A large kinome in a large cell: Stentor coeruleus …...A large kinome in a large cell: Stentor coeruleus possesses highly expanded kinase families and novel domain architectures Sarah - [PDF Document] (2024)

A large kinome in a large cell: Stentor coeruleus possesseshighly expanded kinasefamilies and novel domain architectures

Sarah B. Reiff1 and Wallace F. Marshall1,2,*

1 Department of Biochemistry & Biophysics, University ofCalifornia San Francisco, CA, USA 941432 Center for CellularConstruction, University of California San Francisco, CA, USA94143

* Corresponding author: [emailprotected]

Abstract

Background Stentor coeruleus is a large ciliated protist, about1mm in length, with the extraordinary ability tofully regenerateeach fragment after being cut into pieces, perfectly restoring cellpolarity and morphology. Single-cellregeneration in Stentor remainsone of the greatest long-standing mysteries of biology, but therecently publishedStentor genome now enables studies on thisorganism at the molecular and genetic levels. Here we characterizethecomplete complement of kinases, or kinome, of Stentor, in orderto begin to understand the signaling capacities thatunderlieStentor ’s unique biology.

Results The genome of S. coeruleus contains over 2000 kinases,representing 6% of the predicted proteome.Classification of thekinase genes reveals large expansions in several kinase families,particularly in the CDPKs, theDYRKs, and in several mitotic kinasefamilies including the PLKs, NEKs, and Auroras. The large expansionof theCDPK and DYRK kinase families is an unusual feature of theStentor kinome compared to other ciliates withsequenced genomes.The DYRK family in Stentor, notably, contains only a singlepseudokinase which may suggest animportant role in Stentor growthand survival, while the smaller PEK family contains a novelpseudokinase subfamily.The Stentor kinome also has examples of newdomain architectures that have not been previously observed inotherorganisms.

Conclusion Our analysis provides the first gene-level view intothe signaling capabilities of Stentor and will laythe foundationfor unraveling how this organism can coordinate processes ascomplex as regeneration throughout agiant cell.

Background

Stentor coeruleus stands out for its extremely large sizeandhighly complex behavioral repertoire. The cell is1mm long, with anoral apparatus at the anterior endthat it uses for filter feeding,and precisely arrangedciliary rows that run from the anterior tothe posterior(Figure 1). Despite being a unicellular organism,Stentorhas the ability to fully regenerate after being cut inpiecesin a way that perfectly preserves the original cellpolarity andmorphology [1]. Stentor can also respond tolight and mechanicalstimuli by changing swimmingdirection or by contracting. Thiscontractile response issubject to habituation [2], even though thisphenomenon

is more often associated with multicellular nervoussystems. Thewide range of behaviors suggests a richintracellular signalingcapacity, especially since the cell isalso so large, requiringsignals to propagate over longintracellular distances.

Recently our group has sequenced and assembled adraft genome forS. coeruleus [3], now publiclyavailable(http://stentor.ciliate.org/). The 83 Mb genome has ahighgene density, encoding over 34,000 genes. It is alsoquiteintron-poor, and amazingly the vast majority ofintrons are only15-16 bp in length, making Stentor theorganism with the smallestintrons known to date. Therelease of the genome presents newopportunities for

1

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://stentor.ciliate.org/http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

Figure 1. Stentor coeruleus cell structure. Micro-graph of asingle Stentor cell with a few cortical featureslabeled. Anterior(ant.) and posterior (post.) are indi-cated.

studying the molecular details behind Stentorregeneration andbehavior. How might the complexity ofStentor behavior, habituation,and regeneration bemanifested in terms of the diversity ofsignaling proteinsencoded by the genome?

Protein kinases are ubiquitous throughout biology andplay keyroles in all kinds of cell signaling pathways.Found in all domainsof life, they have achieved a richdiversity through the course ofevolution. Eukaryoticprotein kinases (ePKs) share a common domainfold [4]and are classified into the following 7 main groups, eachofwhich contain several related families and subfamilies:AGC, CAMK,CK1, CMGC, RGC, STE, TK, and TKL.ePKs that aren’t related to any ofthese groups areclassified as “Other” kinases, and includesomewell-known kinases families like Aurora kinases andPolo-likekinases. In addition to ePKs, there are alsoatypical proteinkinases, which are protein kinases thatdon’t share the typicalkinase domain architecturecharacteristic of ePKs. These include thehistidinekinases as well as the RIO, ABC, Alpha, and PIKKfamilies.The evolutionary history of kinase families isclosely coupled tothe evolution of developmental andregulatory processes, such thatthe kinome of anorganism is important for understanding theevolution ofa cell’s behavior.

Stentor belongs to the ciliates, a protist phylumcharacterizedby complex cortical ultrastructure and adiverse range of behaviors.While the kinases of metazoa,Arabidopsis, and many unicellularorganisms tend tomake up around 2-3% of protein coding genes[5–7],ciliates tend to have larger kinomes than mostotherorganisms. The kinases of Tetrahymena and Parameciummake up 4%and 7% of the predicted proteomes,respectively [8, 9]. Theexpansion of kinase families inciliates may suggest a need formultiple distinct butrelated kinase functions in regulating thebehaviors ofthese large and complex free-living cells.

As a first step towards dissecting signaling in Stentor,wecharacterize the kinome of S. coeruleus, with the idea

that most important cellular processes are likely to beregulatedby kinases. We find that the Stentor genomeencodes over 2000kinases, with expansions in severaldifferent kinase families,reflecting the elaborate signalingneeds of this cell.

Materials and Methods

Identification and classification of kinases andotherproteins

Profile HMMs were downloaded for all the kinase groups,families,and subfamilies in Kinbase(http://kinase.com/web/current/kinbase/). These were compared toall theStentor gene models (available at stentor.ciliate.org)usingthe hmmsearch tool in HMMER v3.1 (http://hmmer.org/), using thefollowing command:

hmmsearch -E 0.001 –cpu 16 –noali –seed 544 –tblout

kinases hmmsearch tab.txt -o kinases hmmsearch.txt

kinasedomains.hmm Stentor proteome.fasta

Gene models with partial kinase domains were comparedto the S.coeruleus genomic locus and vegetativeRNA-seq data [3]. In 24 caseswe were able to improvethe gene models, resulting in full kinasedomains, whichhave now been updated on the Stentor genomedatabase(stentor.ciliate.org). All S. coeruleus kinase hitswerethen verified with BlastP [10] using a custom blastdatabase ofall kinase domains in Kinbase. If the highestscoring BlastP hitmatched the highest scoring profileHMM, the gene was classifiedaccordingly, at either thesubfamily, family, or group level. Forthose that didn’tmatch, we used an all-against-all blastp search toclusterthem into either pre-existing kinase families or newStentor-specific kinase families, or as unique. Note thatin KinBase, forthe Tetrahymena genes that are classifiedinto the Ciliate-E2bkinase subfamily, the majority ofthese genes were later reannotatedas PKG kinases in theTetrahymena genome database. Thus, Stentorgenes thatwere matches for the Ciliate-E2b family havebeenclassified as PKG family kinases here. Additionally, sincethepublished kinome analysis of Paramecium [8] onlyclassifieseukaryotic protein kinases at the group level, wehave used thissame approach to classify P. tetraureliakinases into families andsubfamilies as well as to identifyatypical protein kinases(Additional File 1 Table S2), sothat we could make directcomparisons.

To identify other types of proteins, namely cyclins,GPCRs, andguanylate cyclases, we used a similarHMMER/BlastP approach. Forsearches withhmmsearch, profile HMMs were downloaded from Pfamforthe following Pfam domains: PF00134, PF02984, and

2

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://kinase.com/web/current/kinbase/http://kinase.com/web/current/kinbase/http://stentor.ciliate.org/http://hmmer.org/http://hmmer.org/stentor.ciliate.orghttp://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

PF08613 for cyclins; PF00211 for adenylate/guanylatecyclases;PF02116, PF02117, PF02175, PF10192,PF10324, PF10292, PF10316,PF10317, PF10318,PF10320, PF10321, PF10323, PF10326,PF10328,PF12430, PF08395, PF10319, PF10322, PF10325,PF10327,PF14778, PF00001, PF00003, PF00002,PF02076, PF03125, PF01036,PF05462, PF11710,PF02118, PF03383, PF03402 for GPCRs. Thepfamgathering cutoff for each profile HMM was used as thereportingthreshold.

Identification of conserved tyrosinephosphorylation sites

For the MAPK family, CDC2 subfamily, and H2A familyproteins inStentor, sequences were aligned in MEGA7using MUSCLE [11] with theUPGMB clustering methodand default parameters. Sequences were thenchecked forthe following conserved residues in the alignment:Y204from human ERK1 in the MAPK alignment, Y15 from S.pombe CDC2 inthe CDC2 alignment, and Y57 fromhuman H2A in the H2A alignment.

Analysis of domain architecture

Profile HMMs of all the domains in pfam-A weredownloaded fromftp://ftp.ebi.ac.uk/pub/databases/Pfam/releases/Pfam27.0/. Stentorkinase genes werecompared to these with the hmmsearch tool inHMMER,using the pfam gathering cutoff as the reportingthreshold,with the following command:

hmmsearch –cut ga –cpu 16 –noali –seed 544 –domtblout

kinases pfam tab.txt -o kinases pfam.txt Pfam-A.hmm

Stentor kinases.fasta

Phylogenetic analysis

To generate the ePK tree, we removed all the atypicalkinases,leaving only the Stentor ePK sequences, andthen removed genes withtruncated kinase domains (lessthan 100 amino acids). Thesesequences were theninitially aligned to the pfam PKinase domainprofilehmm using the hmmalign tool in HMMER3. Theresultingalignment was trimmed with TrimAl [12] andfurther aligned usingSATe II [13]. The final alignmentwas then used in RAxML v8.2 [14]to generate amaximum likelihood (ML) tree under the WAG aminoacidsubstitution model and gamma model of rateheterogeneity with 200bootstrap replicates. Code foralignment, trimming, and ML analysisis available athttps://github.com/sbreiff/StentorKinome. Theresultingtree (Additional File 2) was visualized usingFigTree(http://tree.bio.ed.ac.uk/software/figtree/).

The Stentor kinase alignment was also used toidentify kinaseswith catalytic site substitions. We lookedfor substitutions in anyof the following three sites fromthe Pkinase domain consensus: K30,D123, D141, whichcorrespond to the lysine and aspartate residues intheconserved motifs VAIK, HRD, and DFG, respectively.

To generate the PEK tree, kinase domains fromhuman, mouse, fly,yeast, Dictyostelium, andTetrahymena sequences were downloaded fromKinbase,and Stentor and Paramecium sequences were aligned totheseusing MUSCLE in MEGA7. The resultingalignment was then trimmed withTrimAl using the“automated1” heuristic. The trimmed alignment wasthenanalyzed in RAxML under the LG substitution modeland gammamodel of rate heterogeneity, with 250bootstrap replicates, usingthe following command:

raxmlHPC-PTHREADS-AVX –T 16 –f a –m

PROTGAMMAAUTO –x 12345 –p 12345 -# autoMRE –s

PEKalign trim.fasta –n PEKtree

The resulting tree was visualized in Figtree. Thesecondpseudokinase domains of metazoan GCN2 were used asan outgroupto root the tree.

Results

The Stentor coeruleus genome containsover 2000 kinases

We used HMMER3 to search for kinase domains in the34,506 Stentorprotein-coding genes. In total weidentified 2057 kinase genes(Additional File 1 Table S1),equaling 6.0% of protein-coding genesin Stentor.Although this is a much higher complement of kinasesthanmost other eukaryotes, it is comparable with otherciliates, whichtend to have higher numbers of kinases(Table 1). Using BlastP toconfirm identities of hits, wealso classified the Stentor kinasesinto kinase groups,families, and subfamilies. While most HMM andBlastPtop hits corresponded, in roughly 10% of cases, the bestHMMhit and best BlastP hit did not match, so weutilized blast-basedclustering to classify these. Some ofthese could be classified intopre-existing families, but 66were unique, and 134 were grouped intoprovisionalStentor -specific families. 10 contained partialkinasedomains that were too small to classify. Maximumlikelihoodphylogenetic analysis recovers groupings basedon ourclassifications (Figure 2).

As kinomes of the ciliates T. thermophila and P.tetraurelia havebeen previously characterized, some ofthe differences between theseand Stentor will behighlighted (Table 2). When comparing kinasegroups(Figure 3A), the biggest difference lies in the smaller

3

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

ftp://ftp.ebi.ac.uk/pub/databases/Pfam/releases/Pfam27.0/ftp://ftp.ebi.ac.uk/pub/databases/Pfam/releases/Pfam27.0/https://github.com/sbreiff/StentorKinomehttp://tree.bio.ed.ac.uk/software/figtree/http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

Table 1. Summary table of kinome sizes.

Organism # of Kinases in Genome Kinases as % of Protein-CodingGeneshom*o sapiens 518 2.7%Caenorhabditis elegans 4382.1%Saccharomyces cerevisiae 113 2.0%Arabidopsis thaliana 10193.7%Dictyostelium discoideum 285 2.6%Phytophthora infestans 3722.1%Giardia intestinalis 278 4.3%Plasmodium falciparum 891.6%Ichthyophthirius multifiliisa 671 8.3%Tetrahymena thermophilaa1068 4.0%Paramecium tetraureliaa 2841 7.2%Stentor coeruleusa 20576.0%a Ciliate.Data sources: H. sapiens [5], C. elegans [15], S.cerevisiae [7], A. thaliana [6], D. discoideum [16], P. infestans[17], G.intestinalis [18], P. falciparum [19], I. multifiliis [20],T. thermophila [9], P. tetraurelia [8]. Note that the originalP.tetraurelia analysis only examined eukaryotic protein kinases,but here we also include atypical protein kinases found inthegenome.

proportion of “Other” kinases in the Stentor kinome, butthis islargely due to the presence of fewer members ofciliate-specificfamilies in Stentor. Stentor also has asmaller proportion ofatypical kinases, but we find severalkinase families that exhibitedsignificant expansions inmembers in Stentor compared to otherorganisms (Figure3B; Table 2). Some of the major kinase families ineachgroup are discussed below.

Atypical Kinases

Stentor encodes 72 genes for Atypical kinases. Becausethesepossess a domain fold significantly different fromthe ePK domain,these could not be included in ourphylogenetic analysis with theother kinases. Among theatypical kinases in Stentor are 3 ABC1kinases, 3 Alphakinases, 3 RIO kinases, 19 PIKKs and 44Histidinekinases (HisKs). Many of these families are poorly

understood, but HisKs are known to mainly functionintwo-component signaling systems and often serve todetectextracellular signals. When the Tetrahymenathermophila macronucleargenome was published, one ofthe notable findings in terms of thekinases encoded wasa large expansion of histidine kinases [9]; theParameciumkinome exhibits a similar pattern (Additional File 1TableS2). While the number of HisKs in Stentor is higher thaninmost other eukaryotes, it is reduced compared toTetrahymena andParamecium (Table 2 and Figure 3B),even though the Tetrahymenakinome is smaller.

AGC Group

Among the ePKs, Stentor contains a total of 348 AGCgroupkinases. This group was named for Protein KinasesA, G, and C, butalso includes the NDR, RSK, andMAST families. Consistent with otherciliates, Stentorencodes no hom*ologs of Protein Kinase C. However,thereare 23 PKAs and 79 PKGs. The presence of multiplePKA and PKGorthologs suggests that Stentor usescAMP and cGMP as upstreamregulators of kinaseactivity. We therefore looked for other membersof cAMPand cGMP signaling pathways in Stentor as well.

Stentor has 67 adenylyl/guanylyl cyclases (AdditionalFile 1Table S3). Cyclase domains are very similarbetween adenylyl andguanylyl cyclases so we wereunable to distinguish between these twotypes of proteins,but we did find two main domain architecturesamongthese 66. 26 of them consisted of a single guanylatecyclase(GC) domain, but 38 possessed a domainarchitecture consisting of aP-type ATPase domainfollowed by two GC domains. Both domainarchitecturesare common in alveolates, the superphylumcontainingciliates, but the latter is alveolate-specific [21].InParamecium, some of these guanylyl cyclases areactivated inresponse to calcium signaling [22,23].

In metazoa as well as other eukaryotes, cAMPsignaling is oftendriven by G-protein coupled receptors(GPCRs), so we searched theStentor gene models forthese domains as well. We found that Stentorencodes 80GPCRs. 66 of these appear similar to the cAMPreceptorfamily of GPCRs originally described in Dictyostelium;

4

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

Ciliate-E2U

RSK

NDR

PKG

PKA

Aurora

PLK

CAMKK

CAMKL

CDPK

NEK

PEK

STE11

STE7

STE20

WNK

DYRK

SRPKGSK

MAPK

CDKCK2

RCK

CDKL

CMGC AGC

CAMK

CK1

TKL

STE

Figure 2. Dendrogram of Stentor ePKs. Shaded regions of the treeindicate kinase groups, and outer ringindicates selected kinasefamilies.

the rest group with the secretin family, the rhodopsinfamily,another rhodopsin-like family, the abscisic acidfamily, a glucosereceptor family, and the ODR4-likefamily of GPCRs (Additional File1 Table S3). All ofthese GPCR families are present in otherciliates exceptfor the ODR4-like family, from which Stentorcontainstwo members.

With 67 cyclases, 80 GPCRs, and 102 combinedPKAs and PKGs, itappears that cAMP and cGMPsignaling pathways are expanded in toto,including boththe kinases themselves and their upstreamregulators,suggesting a capacity for the cell to transducesignals

from a wide variety of inputs. Interestingly, we could notfindany hom*ologs of metazoan G proteins. This isconsistent with otherciliates, however, and since otherparts of the signaling pathwayare present, it is possiblethat Stentor and other ciliates encode aG protein that istoo divergent from metazoan G proteins to be foundbyhom*ology-based approaches.

The NDR kinase family comprises 25 members inStentor. Our grouphas previously shown that the kinaseregulator Mob1 is a polaritymarker and essential forproper growth and regeneration in Stentor[24]. Inmetazoa, Mob1 functions in the Hippo pathway that

5

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

Table 2. Selected kinase families in Stentor compared to othermodel organisms.

Group Family Stentor Tetrahymena Paramecium Yeast HumanAGC PKA23 4 23 3 5

PKC 0 0 0 1 9PKG 79 7 37 0 2NDR 25 5 15 4 4MAST 0 3 19 0 5RSK 485 18 3 8

CAMK CDPK 308 27 115 0 0CAMKL 107 22 96 12 20

CMGC DYRK 142 10 24 1 10CDK 28 15 38 7 21CDKL 14 4 11 0 5CK2 173 16 2 2GSK 18 7 28 4 2RCK 31 8 23 1 3SRPK 15 3 16 1 3MAPK 45 15 686 14

CK1 CK1 130 20 164 4 7STE STE7 5 4 9 4 7

STE11 39 8 25 5 8STE20 29 7 10 7 30

TKL 46 8 15 0 43Other Aurora 44 15 63 1 3

NEK 89 56 115 1 11PLK 49 9 17 1 4ULK 27 73 244 1 5CDC7 0 1 7 11CAMKK 16 9 18 4 2Wee 2 2 4 1 3WNK 10 2 8 0 4PEK 13 4 9 1 4

Atypical HisK 44 95 186 1 0Alpha 3 9 20 0 6PIKK 19 15 34 5 6ABC13 4 6 3 5RIO 3 2 2 2 3

All 2057 1068 2841 113 518

Data sources: Stentor and Paramecium - this study; Tetrahymena,Yeast, Human – Kinbase.

helps regulate cell proliferation and apoptosis. In thispathway,Mob1 interacts with LATS1/2, which are NDRfamily kinases [25]. InStentor these may serve roles inestablishment and maintenance ofpolarity during growthand regeneration, as have been shown forMob1. TheMAST family is another AGC family that bearssequencesimilarities to NDR kinases [26], and while MAST kinasesarepresent in metazoa and other ciliates, we do not findany members inStentor.

CAMK Group

The CAMK group is mainly composed of calcium-andcalmodulin-dependent kinases and totals 561 kinases inStentor,or 27% of the kinome. Calcium signaling isknown to be crucial tocell physiology, particularly inciliates. In Vorticella its role inthe rapid contraction ofthe stalk has been well-studied [27,28] andinParamecium along with contraction it plays a major role

6

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

StentorParamecium

Tetrahymena

Human

AGC

CAMK

CK1

CMGC

TKL

Other

Atypical

STE

RGC

TK

A

B

14%12%10%

8%6%4%

2%0%

CDPK DYRK NEK PLK CDK MAPKHisK ULK

StentorTetrahymenaParameciumHuman

Comparison of Kinase Family Proportions

NDRPKG Aurora

16%

Figure 3. Comparison of kinomes across 4 species. A. Kinasegroups shown as proportions of the totalkinome in S. coeruleus, T.thermophila, P. tetraurelia, and H. sapiens. Size of circle isproportional to kinome size ineach species. B. Bar graph comparisonof 11 selected families in terms of proportion of kinome. Absolutenumberscan be found in Table 2.

in a variety of processes, especially motility andexocytosis[29–31]. In Stentor, calcium flux is involved inoral regenerationas well as contraction and thephotophobic response [32–34]. Severaltypes of kinasescan respond to calcium, including the CDPK(calciumdependent protein kinases), calmodulin dependentproteinkinase (CaMK), and protein kinase C (PKC).Stentor possesses fivekinases in the CaMK1 family, afamily conserved in metazoa. The CDPKfamily, however,appears to have the most dramatic expansion ofanykinase family in Stentor, with 308 members (Figure 3B).The CDPKsconstitute a family of calcium-dependent butcalmodulin-independentprotein kinases that possess EFhands that bind calcium in acalmodulin-like fashion.They were originally discovered in plants[35], where theyare mainly involved in stress and immunesignaling.CDPKs are absent from metazoa and fungi, but are

present in apicomplexan parasites, a sister group totheciliates, where they help regulate a variety ofprocessesincluding motility, host cell invasion, and lifecycletransitions [36]. Most of the organisms that possessCDPKsincluding plants and other alveolates seem tolack PKC, which playsroles in calcium signaling inmetazoa and yeast. Stentor does,however, encodehom*ologs of proteins that function upstream of PKCinhumans, such as phospholipase C and phosphoinositol-3kinase, soperhaps CDPKs are filling this role in Stentorand otherprotists.

The CaMK-like (CAMKL) family kinases, which arerelated to otherCAMKs but are not themselvescalcium-dependent, total 107 members inStentor. 42 ofthese belong to the 5’- AMP-activated kinase(AMPK)subfamily, which are known to be involved inenergyhomeostasis [37]. An additional 28 belong to the

7

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

microtubule affinity-regulating kinase (MARK) subfamily.Inmetazoa, MARKs help regulate microtubuledynamics [38], so thisexpanded family may helpmodulate the extensive microtubule-basedcorticalpatterning of Stentor.

CMGC Group

The CMGC group is named for the CDK, MAPK, GSK,and CDKLfamilies, and contains 317 kinases in Stentor,representing a higherproportion of the kinome thanobserved in other ciliates or inhumans (Figure 3A).TheCDK family contains the canonical cell cycleregulatorsof other systems. Stentor encodes 28 CDKs, including11members of the cdc2 family. Generally, CDKs areregulated bybinding to different cyclins at differentstages of the cell cycle.To determine whether this typeof regulation is also conserved inStentor, we usedHMMER to identify genes with cyclin domainsamongthe predicted proteome, and confirmed hits with BlastP.We finda total of 106 cyclin genes, including 25 thatcontain both theN-terminal and C-terminal cyclindomains (Additional File 1 TableS3). In addition, two ofthe CDK genes possess cyclin domains, ashas beendescribed in Phytophthora infestans [17]. Metazoansystemstypically possess less than 30 different cyclins, so106 representsa massive expansion in Stentor.

Stentor also contains a startling 142 members of thedualspecificity YAK-related kinase (DYRK) family,which has the abilityto phosphorylate tyrosine as well asserine and threonine. Thiscomes to 6.9% of the kinome,compared to 10 members in Tetrahymenaand 24 inParamecium (0.9% of their respective kinomes) and 10inhumans (2.0%). In humans, Dyrk1A is thought tocontribute to DownSyndrome, and Dyrk2 is involved inhedgehog signaling and cell cyclecontrol [39–41]. OtherDYRKs include PRP4 which is involved inmRNAsplicing and Yak which negatively regulates the cell cycleinbudding yeast and modulates G-protein signalingduring growth inDictyostelium [42–44]. Of the StentorDYRK family members, 56 belongto the Dyrk2subfamily, while there are 8 in the plant-likeDyrkPsubfamily, 5 in the Yak subfamily, and 1 PRP4hom*olog,suggesting that multiple DYRK clades contributed to theDYRKfamily expansion in Stentor. The remaining 72weren’t classifiedinto subfamilies, but appear mostsimilar to DYRK2.

Also among the CMGC group in Stentor are 15SRPKs, 17 CK2s, and31 RCKs. SRPKs typicallyphosphorylate SR-rich proteins and playimportant rolesin regulating mRNA maturation [45,46]. CK2sareimplicated in a wide array of cellular functions, havingover 300targets in humans [47,48]. RCKs generally

aren’t well characterized, but the MOK and MAKsubfamilies areimplicated in the regulation ofciliogenesis [49]. In particular,null mutants of theChlamydomonas reinhardtii MOK kinase LF4 resultinabnormally long flagella [50].

Finally, there are also 18 GSKs and 14 CDKLs, aswell as 45MAPKs. MAP kinase signaling cascades arefound throughout eukaryotesand play essential functionsin signal transduction, typicallylinking an extracellularsignal to an alteration in gene expression.The MAPKfamily in Stentor includes 13 members of the ERKsubfamilyand 8 members of the ERK7 subfamily.

STE Group

The STE group is named for the yeast sterile kinases.Stentorcontains 104 STE kinases, which represents agreater proportion ofthe kinome than in other ciliatesbut much lower than metazoa orDictyostelium (Figure3A) [5, 16]. In Stentor 5 STE kinases are inthe STE7family, 39 are in the STE11 family, and 29 are in theSTE20family. In other systems, these families containkinases involved inthe MAPK cascade, namely MAPKkinases (MEKs), MEK kinases (MEKKs),and MEKKkinases (MAP4Ks). The MEKs are subfamilies of theSTE7kinase family and the MEKKs are subfamilies ofthe STE11 kinasefamily, while the MAP4Ks aremembers of the KHS subfamily of theSTE20 kinases.Stentor kinases in the STE7 and STE11 familiesappearto have hom*ology to MEKs and MEKKs by BLAST, andmay servethese roles. However, although Stentor alsocontains STE20 familykinases, these do not seem to beargreat similarity to MAP4Ksspecifically.

Of the Stentor STE20 kinases, 2 belong to the MSTsubfamily.These are orthologs of the Drosophila Hippokinase, which serves toactivate NDR kinases in the Mob1pathway mentioned earlier. InTetrahymena, an MSThom*olog has been shown to be important forproperplacement of the cell division plane [51].

CK1 and TKL Groups

Consistent with other ciliates, Stentor does not encodeanykinases from the RGC or TK group, which consist ofreceptor guanylylcyclases and tyrosine kinases,respectively. Indeed, to this pointthese groups havemainly been observed in metazoa andchoanoflagellates.Stentor does, however, contain 46 TKL (TK-like)kinases,which have a similar domain fold to the TKsbutphosphorylate on serine/threonine. While this group islarger inStentor than other ciliates, it represents a muchsmaller proportionof the kinome than in metazoa orDictyostelium (Figure 2A) [5, 16].There are also 130

8

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

members of the CK1 group, which consists only of thecaseinkinase 1 family in Stentor.

“Other” Kinases

These comprise kinase families that aren’t related to anyof theprevious groups, and total 460 kinases in Stentor.This includessome major mitotic kinase families, such asAurora kinases, PLKs,and NEKs. The aurora kinasesare well studied for their roles inprogression throughmitosis, and Stentor encodes 44 aurora kinasegenes. Inaddition, Stentor also contains 49 PLKs and 89 NEKs.Thesefamilies are known to play roles in both the cellcycle and inciliogenesis in metazoa. With a highlypatterned cell cortexcontaining tens of thousands of cilia,expansions in these familiesmay allow Stentor to keep upwith the ciliogenesis needs of the cellthroughout the cellcycle. Along with the previously mentioned NDRsandCDKs, mitotic kinases in Stentor often seem to exhibitexpandedfamily sizes (Figure 3B).

A notable finding in the Tetrahymena kinome was alarge expansionof the ULK kinases [9], which is also trueof Paramecium (Figure 3B,Table 2). The functions ofthe ULK kinases are not very wellunderstood, though inother systems they are reported to be involvedin avariety of activities including hedgehog signaling,motileciliogenesis, autophagy, and ER-to-Golgitrafficking [52–54].In contrast to the other ciliatesmentioned, Stentor only contains27 ULKs, less than halfthe number found in Tetrahymena.

The kinase analysis in the Tetrahymena genomedelineated severalputative ciliate-specific kinasefamilies [9]. Many of thesefamilies appear not to beconserved in Stentor, but we do find 7kinases in theCiliate-A9 family, 11 in the Ciliate-C1 family, 3 intheCiliate-C4 family, and 52 in theCiliate-E2-Unclassifiedsubfamily (Additional File 1 Table S1). The“Other”kinases in Stentor also include 10 WNK kinases, 16CAMKKs, 13PEKs, and 2 Wee kinases.

Tyrosine Phosphorylation and DualSpecificity Kinases

While Stentor lacks canonical TKs, other organismslacking TKs docontain phosphotyrosine, usuallyproduced by cryptic tyrosinekinases or dual-specificitykinases from other kinase groups. Abovewe described alarge expansion in DYRKs in Stentor. However,whilepreviously studied DYRKs autophosphorylate ontyrosineresidues, they only transphosphorylate on serineorthreonine residues [55]. Thus, this family is not expectedtodrive tyrosine phosphorylation of other substrates inStentor, eventhough they are highly expanded.

Other kinase families have been shown to exhibit dualspecificityin their transphosphorylation activities,including STE7, Wee, andCK2. In yeast and metazoa,Wee1 phosphorylates the tyrosine residueof the motifGxGTYG on CDK1 (CDC2), leading toinactivation [56,57].Stentor has 11 kinases in the CDC2subfamily, so we performed analignment and looked forthis particular motif. This motif wasconserved in 10 ofthe CDC2 kinases (Figure 4A), suggesting thatthesubstrate for the Stentor Wee kinases is conserved, andlikelyits tyrosine phosphorylation activity as well. Theonly CDC2 kinasethat doesn’t conserve this sequence isdivergent across the wholekinase domain including atpositions of critical catalytic sites,and happens to be oneof the CDKs that possess two cyclin domains attheC-terminus (Figure 4B), raising the possibility that itmay actprimarily as a cyclin that uses a pseudokinasedomain as ascaffold.

The STE7 family, also known for dualspecificitytransphosphorylation, consists of MEKs (MAPkinasekinases). It is known in metazoa that MAPKs areactivated whenthe threonine and tyrosine residues in aTxY motif in the activationloop arephosphorylated [58,59]. In ERK1/2, this sequenceistypically TEY, but in other MAPK subfamilies themiddle residuecan be different [58,59]. We aligned allthe Stentor MAPKs andlooked for conservation of thismotif. The TxY activation loop wasfound to be presentin 31 of the 45 Stentor MAPKs, most commonly asTEYor TDY. This number includes 12 of 13 ERK subfamilykinases andall 8 ERK7 subfamily kinases (Figure 4C).Conservation of a majorSTE7 kinase substrate suggestsconserved tyrosine phosphorylationability.

While the function of Wee kinases and MEKs havebeen wellstudied, CK2 kinases are much more pleiotropicand not as wellunderstood. They are ubiquitous,constitutively expressed, andthought to act on hundredsof substrates, many of which seem to beinvolved intranscriptional activity and the spliceosome [48,60].Formost substrates, CK2 probably phosphorylates on serineorthreonine residues, but recently CK2 was found tophosphorylatehistone H2A on Y57 to help regulatetranscriptional elongation [61].We searched the Stentorpredicted proteome for hom*ologs of H2A andidentified21 proteins with an H2A domain. We aligned these tothehuman H2A protein to look for conservation of Y57,and found that 20of the 21 Stentor proteins had thisresidue conserved, suggestingthat the tyrosinephosphorylation ability of CK2 is conserved inStentor(Figure 4D). We thus infer from this analysis of kinaseandsubstrate sequences that Stentor is likely to usetyrosinephosphorylation of substrates as a regulatorymechanism, relying ondual specificity kinases.

9

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

LTEYVATRWYRAPEIMLLTEYVVTRWYRAPEVVLLTEYVVTRWYRAPEVVLLTEYVVTRWYRAPEVILLTEYVVTRWYRAPEVVLLTEYVVTRWYRAPEVVLLTEYVVTRWYRAPEVILLTEYVVTRWYRAPEVILYTDYVVNRWYRAPEILLLTEYVVTRWYRAPEVILIEDYVNNRWYRAPELILLTEYVVTRWYRAPEVVLLTEYVVTRWYRAPEVILLTEYVVTRWYRAPEVVLVTEYVATRWYRAPEVLLLTDYVATRWYRAPEILLLTDYVATRWYRAPEILLLTDYIASRWYRAPEVLLLTDYVASRWYRAPEVLLLTDYVATRWYRAPEILLLTDYVATRWYRAPEILLLTDYVATRWYRAPEILLLTDYVATRWYRAPEILL

TxYHsERK1 201 SteCoe_11125 183SteCoe_1118 183SteCoe_13752198SteCoe_16181 183SteCoe_17068 183SteCoe_18200 183SteCoe_18417198SteCoe_22275 179SteCoe_23681 198SteCoe_28331 182SteCoe_28657183SteCoe_31654 198SteCoe_8174 183HsERK7 174SteCoe_14585172SteCoe_20709 172SteCoe_23361 171SteCoe_35001 171SteCoe_37910172SteCoe_4317 172SteCoe_518 172SteCoe_6250 172

*ScereH2A 51HsH2A 50SteCoe_25962 54SteCoe_14254 54SteCoe_3104654SteCoe_30646 54SteCoe_27208 54SteCoe_5261 54SteCoe_2531354SteCoe_32925 54SteCoe_13935 59SteCoe_35065 61SteCoe_2160756SteCoe_30304 56SteCoe_9474 56SteCoe_3078 56SteCoe_970266SteCoe_16276 64SteCoe_26943 62SteCoe_25404 57SteCoe_285257SteCoe_16345 41SteCoe_30853 500

VYLTAVLEYLAAEILEVYLAAVLEYLTAEILEVYLAAVLEYLAAEVLEVYLAAVLEYLAAEVLEVYLAAVLEYLAAEVLEVYLAAVLEYLAAEVLEVYLAAVLEYLAAEVLEVYLAAVLEYLAAEVLEVYLAAVLEYLAAEVLEVYLAAVLEYLAAEVLEVYLSAVLEYLAAEVLEVYAGAVLEYLTAEVLEVYTGAILEYLTAEVLEVYTGAILEYLTAEVLEVYTGAILEYLTAEVLEVYAGAILEYLTAEVLEIGVSAVLEYLTAEILEIGVAAVLEYLAAEVLEVSICAVLEYLTTEILDVTVAAVLEYLTAEVIEVTVAAILEYLTAEVIEVFLTAVLEFMSRELLTIFLAGTLEYLCAEVLD

GxGTYGIGEGTYGVVYKARHKIGEGTYGVVYKGRHKIGTGTYGVVYKAKDKIGEGTYGVVYKARDTIGEGTYGVVYKSRDKIGTGTYGVVYKAKDKVERRSKGKVYTGMLRIGTGTYGVVYKAKDKIGEGTYGVVYKTLDKIGEGTYGVVYKSRDKIGEGTYGVVYKSRDKIGEGTYGVVYKSRDKIGEGTYGVVYKSRDK

SpombeCdc2 10HsCDK1 10 SteCoe_11769 17SteCoe_1204140SteCoe_19115 15SteCoe_24201 17SteCoe_2757 108SteCoe_3003817SteCoe_32943 14SteCoe_3497 15SteCoe_5190 15SteCoe_570515SteCoe_7783 15

CDK

CDK Cyc_N Cyc_CF-box

Canonical

SteCoe_2757

A B

C D

Figure 4. Conserved tyrosine phosphorylation sites in Stentorproteins. A: Alignment of S. pombe cdc2and H. sapiens CDK1 withStentor hom*ologs at the GxGTYG tyrosine phosphorylation motif. B:Comparison ofcanonical cdc2 and SteCoe 2757 domain architectures.Drawn to scale. C: Alignment of H. sapiens ERK1 and ERK7sequenceswith Stentor hom*ologs at the TxY phosphorylation motif. D:Alignment of S. cerevisiae and H. sapiensH2A sequences with Stentorhom*ologs at the Y57 phosphorylation site.

10

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

Table 3. Other protein domains present on Stentor kinases.

Domain # Domain # Domain # Domain #EF-hand 287 FYVE 6 ADK 2 DEP1cNMP binding 97 HEAT 5 Cyclin C 2 DUF547 1Ankyrin 93 LRR 5 CyclinN 2 DUF3543 1POLO box 43 FAT 4 DUF3385 2 NUC194 1Response regulator43 WD40 4 F-box 2 RCC1 1HATPase C 42 FATC 3 IFT57 2 Ribonuclease2-5A 1TPR 14 Myosin TH1 3 Rapamycin binding 2 ubiquitin 1zf-MYND 9NAF 3 RWD 2 VWA 1PH 8 PG binding 3 ANTH 1

Novel domain architectures in theStentor kinome

Highly expanded kinase families create potentialfordiversification and subfunctionalization in its members.Genefusions could potentially result in acquisition ofadditionalprotein domains that could facilitate greaterfunctional diversitywithin a kinase family. We usedHMMER3 to search the Stentor kinasesfor additionalpfam domains. Many of the protein domains uncoveredinthe kinases helped confirm classifications: the polo-likekinasehom*ologs had polo domains, for example, whilePKGs hadnucleotide-binding domains and CDPKspossessed EF hands for calciumbinding (Table 3).However, we also uncovered examples of noveldomainarchitectures that have not been previously described inothersystems (Figure 5). One such architecture is foundon six Stentorkinases (SteCoe 19985, SteCoe 20512,SteCoe 31546, SteCoe 35363,SteCoe 7134, andSteCoe 7398) and consists of a PAKA kinasedomainwith a FYVE domain near the N-terminus (Figure 5B).FYVEdomains generally serve to target a protein intoendosomalmembranes, and have been observed onkinases of other familiesbefore [17], but this is the first

example of a STE kinase with a FYVE domain. A morestrikingdomain architecture, not previously described forany kinase, isfound on two CDPKs (SteCoe 9200 andSteCoe 27913) with an adenylatekinase domain at theN-terminus (Figure 5A). Like other CDPKs, theyalsohave EF hands at the C-terminus, and SteCoe 9200possesses afifth EF hand between the adenylate kinaseand protein kinasedomains.

We also identified two NEK kinases (SteCoe 18659and SteCoe22983) with an IFT57 domain (Figure 5C).IFT57 is a protein involvedin intraflagellar transport(IFT) [62]. It helps stabilize the IFT-Bcomplex [63,64],and the human hom*olog HIPPI is involved intheregulation of apoptosis, as well as cilia assembly andsonichedgehog signaling [65,66]. In most systems IFT57domains are notfound on kinases, though the domainarchitecture found in Stentor isalso found in otherciliates. Perhaps the IFT57-like kinases ofciliates mayprovide more clues about the role of IFT57 in IFTandits interactions with other proteins.

Similarly, seven of the Aurora kinases in Stentorpossess anN-terminal zf-MYND domain, which canmediate protein-proteininteractions. While this domainarchitecture has not been observedon Aurora kinases

EF handsA

BPAKA Kinase

FYVENEK Kinase IFT57

C

ADK CDPK 1 1805

1 1428 681

Figure 5. Novel domain architectures among Stentor kinases. A:Domain architecture of SteCoe 9200, aCDPK family kinase with anadenylate kinase domain at the N-terminus. Like most CDPKs, thereare also EF handsat the C-terminus. B: Domain architecture ofSteCoe 7134, a PAKA kinase with an N-terminal FYVE domain. C:Domainarchitecture of SteCoe 18659, an NEK family kinase with an IFT57domain at the C-terminus. Numbersrefer to amino acid positions.Drawn to scale.

11

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

from other eukaryotes, it is found in other ciliates.

Pseudokinases

Active kinases belonging to the ePK superfamilygenerally containtwelve conserved subdomains [4], butmost eukaryotic proteomes alsoencode pseudokinases.These are proteins with domains that arehighly similarto kinase domains but with one or more criticalresiduein the catalytic site altered [67]. This usually results inalack of kinase activity, although there are cases ofpredictedpseudokinases that have been experimentallyshown to have some levelof kinase activity [68–71].However, even when a pseudokinase isknown to becatalytically inactive, often the protein is stillimportantto cell physiology, for instance in allosteric regulationofan active kinase or in scaffolding of an enzymecomplex[67,72,73]. Substitutions in the lysine andaspartate residues ofthe active site motifs VAIK, HRD,and DFG are often used to predictpseudokinases, andstudies suggest that pseudokinases tend to makeupabout 10% of the kinome in humans, about 11% inDictyostelium, andabout 22% in Tetrahymena andParamecium [5, 8, 16]. We searchedStentor kinases forchanges in one of these three highly conservedresidues.The atypical protein kinases were not included inthissearch, since their kinase domains are significantlydifferentfrom the well-defined ePK protein kinasedomain fold. Of the 1977Stentor ePKs, 289 (14.6%)were found to have a substitution in oneof these threeresidues. Twelve of these were hom*ologs of Scyl orWnk.Scyl is a family found across eukaryotes composedentirely ofpseudokinases; Wnk kinases are generallyactive but lack theconserved lysine residue [67].

We further examined whether any individual kinasefamilies hadhigher rates of active site substitutions inStentor (Figure 6A). Alarge number of thepseudokinases were unique or in Stentor-specific families(39 and 47, respectively), indicating a highdegree ofdivergence across other sites in these genes.

Of the kinases from conserved families, we found thatthe CDPKfamily in Stentor included 42 predictedpseudokinases out of 308(13.6%) and the CK1 familyincluded 23 out of 130 (17.7%). We notethat the Giardiakinome contains 195 NEKs, a total of 70% of itskinome,and a surprising 71% of these were identifiedaspseudokinases [18]. Thus, there is a precedent formembers oflarge kinase families evolving intopseudokinases, and perhaps thesehelp regulate the activekinases.

Notably, we also found that the PEK family inStentor contained10 pseudokinases, representing 77% ofthe family. In the RNA-seqdata from vegetative cells

that helped inform the original gene models [3], thesePEKpseudokinases have a mean of 1691 reads mappingcompared to mean of947 for the active PEKs (AdditionalFile 1 Table S1), suggestingthat they are expressedduring vegetative growth and are notpseudogenes.Furthermore, maximum likelihood analysis suggeststhatthey group together in a single subfamily (Figure 6B). Inplaceof the normally-conserved DFG motif, we find mostof these sequenceseither possess a DYD or NFR motif.Interestingly, sequencecomparison between this subfamilyand active PEKs reveals that whilethe HRDLKPxN andDFG motifs aren’t well conserved in this subfamily,theNIFLD motif in between is slightly more so (Figure 6C).

In contrast, other kinase families in Stentor exhibitedmuchlower numbers of pseudokinases than the 14.6%average for thekinome. In particular, the DYRK familycomprises an impressive 142members in Stentor, butonly one of these is a predictedpseudokinase.Additionally, several families contained nopredictedpseudokinases at all (Figure 6A).

Discussion

Here we have presented the kinome of Stentor coeruleus.We haveidentified 2057 protein kinases, and classifiedthem into groups,families, and subfamilies. Manywell-known kinases from othereukaryotes are conservedin Stentor, including MAPKs, PKAs and PKGs,as wellas several mitotic kinase families including the NDRs,NEKs,PLKs, CDKs, and Auroras. We found manysimilarities to kinomes ofciliates and other protists,including the absence of PKC and the TKkinase group,and an expanded family of CDPKs.

In classical studies of Stentor, two properties tend tostandout: 1) its large size, and 2) its impressiveregeneration ability.What signaling capabilities would bepresent in such a cell?

With respect to cell size, we note that generally theorganismsthat have kinomes of over 1000, mainly plantsand ciliates, alsohave relatively large cells. Polyploidy inthese organisms helpssolve the need for larger scaleprotein production, but it is stilleasy to imagine thatlarger intracellular volumes require richer andmorecomplex signaling repertoires. For example, Stentorhasextensive and complicated cortical patterning. This istrue ofciliates in general more than other phyla, butStentor ’s corticalpatterning is more complex even thanmany other ciliates, especiallywithin the oral apparatus.Oral regeneration alone requires the denovo assembly oftens of thousands of centrioles, withsubsequentciliogenesis. Accordingly, we find that some of thekinasefamilies known to be involved in centriole assembly and

12

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

0% 20% 40% 60% 80%

ULKRCK

STE11STE20

CAMKKPKA

CDKLDYRK

CAMKLRSK

NDRAurora

GSKCK2PLK

MAPKNEKCDKPKG

CiliateSRPKCDPK

CK1STE7

Stentor-sp.Unique

PEK

A B Predicted Pseudokinases in Stentor Kinase Families

GCN2 Subfamily

Stentor PEK Pseudokinase Subfamily

Subdomain VIb Subdomain VIIC

3947

1

1

11

11

00000

2

2

2

2342

3

97

6

3

10

9

Sten

tor P

seud

okin

ases

GCN2

Pt_GSPATP00001794001

Pt_GSPATP00025893001

SteCoe_36879

Tt_6287Dm_PEK.PEK

Mm_GCN2.kindom2

SteCoe_9908

Pt_GSPATP00030527001

Mm_PEK.PEK

SteCoe_34149

SteCoe_2709

SteCoe_14326

SteCoe_4249

Sc_GCN2

Hs_HRI

Hs_GCN2.kindom2

SteCoe_31483

Pt_GSPATP00016393001

Hs_PKR

SteCoe_8747

Dm_Gcn2

SteCoe_33848

Mm_PKR

SteCoe_33517

Tt_18670

Pt_GSPATP00025128001

Pt_GSPATP00012376001

Mm_HRI

Pt_GSPATP00005295001

SteCoe_27790

Mm_GCN2

Hs_PEK.PEK

Tt_11434

Pt_GSPATP00036445001

SteCoe_12670

Tt_21870

SteCoe_22867

Hs_GCN2

Dm_Gcn2.kindom2HRI

PEK subfamily

PKR

NFR

DYD

Figure 6. Pseudokinases among Stentor families. A: Proportionsof pseudokinases are shown as a percentageof total family membersin Stentor. Numbers next to each bar indicate absolute number ofpseudokinases for thatfamily. “Stentor-sp.” refers to Stentor-specific families, “Ciliate” refers to ciliate-specific families,and “Unique” refersto unique kinases in Stentor. B: Maximumlikelihood tree of PEK family kinases in ciliates and metazoa. Redlinesnext to Stentor pseudokinases indicate unusual motifs thatreplace the usual DFG motif. Species represented asfollows: SteCoe– Stentor ; Tt – T. thermophila; Pt – P. tetraurelia; Sc – S.cerevisiae; Hs – H. sapiens; Mm – M.musculus; Dm – D. melanogaster.C: Comparison of HMM logos at kinase subdomains VIb and VII betweentheStentor PEK pseudokinases and the similar GCN2 subfamily. Redline indicates position of motifs indicated on tree.

ciliogenesis in metazoa, like PLKs and NEKs, are quiteexpandedin Stentor. It remains unclear whether themembers in these familiesare functionally redundant, orwhether they have evolved to exhibitgreater specificity in

localization or substrate.Regarding regeneration, particularlyof the oral

apparatus, the process proceeds much like oraldevelopment duringcell division. The cell division

13

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

process even under normal conditions requires manysteps, withformation of an entire new oral apparatus atthe midpoint of thecell, where the anterior end of theposterior daughter needs towiden while the posterior endof the anterior daughter mustconstrict. The steps of celldivision must be tightly regulated, andaccordinglyStentor possesses many cell division kinases, including28CDKs and 44 Auroras.

Calcium signaling is also thought to play a role inregeneration[33]. In ciliates, calcium channels are presentin the plasmamembrane as well as intracellularstructures like the contractilevacuole and the alveolarsacs, and calcium signaling has establishedroles incontraction and motility [29,32, 74]. Furthermore,calciumis known to play a vital role in wound healing inmetazoa [75–78].It produces a clotting reaction inStentor and has been hypothesizedto play a central rolein wound healing after cell cutting [79].While otherciliates contain sizable CDPK families, theStentorgenome contains an impressive 308 CDPKs, making thisthelargest kinase family in Stentor, as well as the largestCDPK familydescribed in an organism so far. LargeCDPK families in ciliates ingeneral may grant these largeand complex cells greater versatilityin calcium signalingcapabilities. The need for this versatility maybe evengreater in Stentor, taking also into consideration itsvastsurface area and impressive wound healing capabilities.Two ofthe Stentor CDPKs even possess N-terminaladenylate kinase domains.This Stentor -specificinnovation could serve to couple energyhomeostasis withcalcium signaling and highlights the potentialforincreased signaling versatility in large families.

Another kinase family known to be involved inregulating celldivision and growth are the DYRK kinases.Stentor possesses 142 DYRKkinases, which is several foldhigher than what is found in otherciliate genomes andappears to be the highest number of DYRKsencoded ina single genome to date. Furthermore, out of all ofthese,only a single one lacks one of the conservedresiduespreviously shown to predict pseudokinases. This findingwasstriking, but previously studied DYRK kinases havebeen shown toself-regulate by autophosphorylation.Thus, it is likely there isselective pressure to retain afunctional active site in order tomaintain regulation.

Stentor has reductions in certain kinase familiescompared toother ciliates, though the reasons why areunclear. In particular,Stentor has fewer members of theULK and HisK families thanTetrahymena orParamecium, but even in these species the roles oftheseexpanded families are not known. However, while theseciliatestend to have a lot in common with Stentor interms of ultrastructureand lifestyle, there are important

differences that set Stentor apart. In addition toStentor ’smuch greater size, there are also behavioraldisparities. Forexample, while Tetrahymena andParamecium generally spend all theirtime swimming,Stentor possesses a sessile state and can easilytransitionbetween sessile and swimming behavior. As it spendslesstime swimming, perhaps Stentor has less need for vastquantitiesof kinases in families like the HisKs, which arevery often involvedin sensing extracellular environments.

Conclusion

A kinome of over 2000 in Stentor underscores a richsignalingcapacity which allows this single cell to maintainits large size,employ a variety of different behaviors, andeven to regenerateafter gross injury. It is characterizedby multiple familyexpansions and contains bothconserved and novel domainarchitectures, representingfunctional conservation as well asinnovation within thekinases and distinguishing it even from otherciliates.

Abbreviations

ePK – eukaryotic protein kinase; CaM – calmodulin;GC – guanylatecyclase; GPCR – G protein coupledreceptor; HisK – histidine kinase;HMM – hiddenmarkov model; IFT – intraflagellar transport; ML-maximum likelihood; TK – tyrosine kinase

Additional Files

Additional File 1. Table S1. Classification of Stentorkinases.Table S2. Further classification of P. tetraureliakinases. TableS3. Stentor non-kinase genes identifiedin this study: GPCRs,cyclins, and guanylyl cyclases.

Additional File 2. Nexus tree file of Stentor ePKMaximumLikelihood analysis.

Acknowledgments

We thank Naomi Stover for hosting the Stentor genomedatabase andfor incorporating our gene modelimprovements. We thank Lydia Brightand TatyanaMakushok for critical reading of the manuscript, andforother members of the Marshall Lab for helpfuldiscussions andcomments.

Funding

We acknowledge support from NIH grant GM113602.

14

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

Availability of Data and Materials

All relevant data are published within the paper anditssupporting additional files. Additionally, Stentor kinasegeneannotations and updated gene models will be visibleon StentorDB(stentor.ciliate.org).

Authors’ Contributions

SBR and WFM conceived and designed the study. SBRanalyzed data,and wrote the paper with help fromWFM.

Competing Interests

The authors declare no conflict of interest.

References

1. Tartar, V.: The Biology of Stentor. InternationalSeries ofMonographs on Pure and Applied Biology,vol. 5, p. 413. PergamonPress, Oxford (1961)

2. Wood, D.C.: Habituation in stentor: aresponse-dependentprocess. J Neurosci 8(7),2248–53 (1988)

3. Slabodnick, M.M., Ruby, J.G., Reiff, S.B., Swart,E.C., Gosai,S., Prabakaran, S., Witkowska, E.,Larue, G.E., Fisher, S., Freeman,R.M.,Gunawardena, J., Chu, W., Stover, N.A., Gregory,B.D., Nowacki,M., Derisi, J., Roy, S.W., Marshall,W.F., Sood, P.: Themacronuclear genome ofstentor coeruleus reveals tiny introns in agiant cell.Curr Biol (2017). doi:10.1016/j.cub.2016.12.057

4. Hanks, S.K., Quinn, A.M., Hunter, T.: Theprotein kinasefamily: conserved features anddeduced phylogeny of the catalyticdomains.Science 241(4861), 42–52 (1988)

5. Manning, G., Whyte, D.B., Martinez, R., Hunter,T.,Sudarsanam, S.: The protein kinasecomplement of the human genome.Science298(5600), 1912–1934 (2002).doi:10.1126/science.1075762

6. Zulawski, M., Schulze, G., Braginets, R.,Hartmann, S.,Schulze, W.X.: The arabidopsiskinome: phylogeny and evolutionaryinsights intofunctional diversification. BMC Genomics 15,548(2014). doi:10.1186/1471-2164-15-548

7. Hunter, T., Plowman, G.D.: The protein kinases ofbuddingyeast: six score and more. TrendsBiochem Sci 22(1), 18–22(1997)

8. Bemm, F., Schwarz, R., Forster, F., Schultz, J.: Akinome of2600 in the ciliate parameciumtetraurelia. FEBS Lett 583(22),3589–92 (2009)

9. Eisen, J.A., Coyne, R.S., Wu, M., Wu, D.,Thiagarajan, M.,Wortman, J.R., Badger, J.H.,Ren, Q., Amedeo, P., Jones, K.M.,Tallon, L.J.,Delcher, A.L., Salzberg, S.L., Silva, J.C., Haas,B.J.,Majoros, W.H., Farzad, M., Carlton, J.M.,Smith, R.K., Garg, J.,Pearlman, R.E., Karrer,K.M., Sun, L., Manning, G., Elde, N.C.,Turkewitz,A.P., Asai, D.J., Wilkes, D.E., Wang, Y., Cai,H.,Collins, K., Stewart, B.A., Lee, S.R., Wilamowska,K., Weinberg,Z., Ruzzo, W.L., Wloga, D., Gaertig,J., Frankel, J., Tsao, C.C.,Gorovsky, M.A.,Keeling, P.J., Waller, R.F., Patron, N.J.,Cherry,J.M., Stover, N.A., Krieger, C.J., del Toro, C.,Ryder, H.F.,Williamson, S.C., Barbeau, R.A.,Hamilton, E.P., Orias, E.:Macronuclear genomesequence of the ciliate tetrahymena thermophila,amodel eukaryote. PLoS Biology 4(9), 286(2006).doi:10.1371/journal.pbio.0040286

10. Altschul, S.F., Gish, W., Miller, W., Myers, E.W.,Lipman,D.J.: Basic local alignment search tool. JMol Biol 215(3), 403–10(1990).doi:10.1016/S0022-2836(05)80360-2

11. Edgar, R.: Muscle: multiple sequence alignmentwith highaccuracy and high throughput. NucleicAcids Research 32(5),1792–1797 (2004).doi:10.1093/nar/gkh340

12. Capella-Gutiérrez, S., Silla-Mart́ınez, J.M.,Gabaldón, T.:trimal: a tool for automatedalignment trimming in large-scalephylogeneticanalyses. Bioinformatics 25(15), 1972–3(2009).doi:10.1093/bioinformatics/btp348

13. Liu, K., Warnow, T.J., Holder, M.T., Nelesen,S.M., Yu, J.,Stamatakis, A.P., Linder, C.R.:Sate-ii: very fast and accuratesimultaneousestimation of multiple sequence alignmentsandphylogenetic trees. Syst Biol 61(1), 90–106(2012).doi:10.1093/sysbio/syr095

14. Stamatakis, A.: Raxml version 8: a tool forphylogeneticanalysis and post-analysis of largephylogenies. Bioinformatics30(9), 1312–3 (2014).doi:10.1093/bioinformatics/btu033

15. Manning, G.: Genomic overbiew of protein kinases.WormBook15, 1–19 (2005).doi:10.1895/wormbook.1.60.1

15

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

16. Goldberg, J.M., Manning, G., Liu, A., Fey, P.,Pilcher, K.E.,Xu, Y., Smith, J.L.: Thedictyostelium kinome–analysis of theproteinkinases from a simple model organism. PLoS Genet2(3), 38(2006). doi:10.1371/journal.pgen.0020038

17. Judelson, H.S., Ah-Fong, A.M.: The kinome ofphytophthorainfestans reveals oomycete-specificinnovations and links to othertaxonomic groups.BMC Genomics 11, 700(2010).doi:10.1186/1471-2164-11-700

18. Manning, G., Reiner, D.S., Lauwaet, T., Dacre, M.,Smith, A.,Zhai, Y., Svard, S., Gillin, F.D.: Theminimal kinome of giardialamblia illuminatesearly kinase evolution and unique parasitebiology.Genome Biol 12(7), 66(2011).doi:10.1186/gb-2011-12-7-r66

19. Ward, P., Equinet, L., Packer, J., Doerig, C.:Proteinkinases of the human malaria parasiteplasmodium falciparum: thekinome of a divergenteukaryote. BMC Genomics 5, 79(2004).doi:10.1186/1471-2164-5-79

20. Coyne, R.S., Hannick, L., Shanmugam, D.,Hostetler, J.B.,Brami, D., Joardar, V.S., Johnson,J., Radune, D., Singh, I.,Badger, J.H., Kumar, U.,Saier, M., Wang, Y., Cai, H., Gu, J.,Mather,M.W., Vaidya, A.B., Wilkes, D.E., Rajagopalan,V., Asai,D.J., Pearson, C.G., Findly, R.C.,Dickerson, H.W., Wu, M., Martens,C., Van dePeer, Y., Roos, D.S., Cassidy-Hanley, D.M., Clark,T.G.:Comparative genomics of the pathogenicciliate ichthyophthiriusmultifiliis, its free-livingrelatives and a host species provideinsights intoadoption of a parasitic lifestyle and prospectsfordisease control. Genome Biol 12(10), 100(2011).doi:10.1186/gb-2011-12-10-r100

21. Linder, J.U., Schultz, J.E.: Guanylyl cyclases inunicellularorganisms. Mol Cell Biochem 230(1-2),149–58 (2002)

22. SCHULTZ, J., POHL, T., KLUMPP, S.:Voltage-gated ca-2+ entryinto paramecium linkedto intraciliary increase in cyclic-gmp.Nature322(6076), 271–273 (1986). doi:10.1038/322271a0

23. Klumpp, S., Schultz, J.E.: Characterization ofaca2+-dependent guanylate cyclase in the excitableciliary membranefrom paramecium. Eur JBiochem 124(2), 317–24 (1982)

24. Slabodnick, M.M., Ruby, J.G., Dunn, J.G.,Feldman, J.L.,DeRisi, J.L., Marshall, W.F.: The

kinase regulator mob1 acts as a patterning proteinfor stentormorphogenesis. PLoS Biology 12(5),1001861 (2014).doi:10.1371/journal.pbio.1001861

25. Hergovich, A.: Mob control: reviewing a conservedfamily ofkinase regulators. Cell Signal 23(9),1433–40 (2011)

26. Pearce, L.R., Komander, D., Alessi, D.R.: Thenuts and boltsof agc protein kinases. Nat Rev MolCell Biol 11(1), 9–22 (2010).doi:10.1038/nrm2822

27. Katoh, K., Naitoh, Y.: Control of cellularcontraction bycalcium in vorticella. J Exp Biol189(1), 163–77 (1994)

28. Misra, G., Dickinson, R.B., Ladd, A.J.: Mechanicsofvorticella contraction. Biophys J 98(12),2923–32 (2010).doi:10.1016/j.bpj.2010.03.023

29. Nakaoka, Y., Tanaka, H., Oosawa, F.:Ca2+-dependentregulation of beat frequency ofcilia in paramecium. J Cell Sci 65,223–31 (1984)

30. Matt, H., Bilinski, M., Plattner, H.:Adenosinetriphosphate,calcium and temperaturerequirements for the final steps ofexocytosis inparamecium cells. J Cell Sci 32, 67–86 (1978)

31. Kerboeuf, D., Le Berre, A., Dedieu, J.C., Cohen,J.:Calmodulin is essential for assembling linksnecessary forexocytotic membrane fusion inparamecium. EMBO J 12(9), 3385–90(1993)

32. Huang, B., Pitelka, D.R.: The contractile processin theciliate, stentor coeruleus. i. the role ofmicrotubules andfilaments. J Cell Biol 57(3),704–28 (1973)

33. Maloney, M.S., Walsh, P.R., Rhoadarmer, K.B.:Calcium andcalmodulin antagonists delay oralregeneration in the ciliatestentor coeruleus.Journal of Experimental Biology 161(1),135–149(1991)

34. Kim, I.H., Prusti, R.K., Song, P.S., Hader, D.P.,Hader, M.:Phototaxis and photophobic responsesin stentor coeruleus. actionspectrum and role ofca2+ fluxes. Biochim Biophys Acta799(3),298–304 (1984)

35. Olah, Z., Bogre, L., Lehel, C., Farago, A., Seprodi,J.,Dudits, D.: The phosphorylation site ofca(2+)-dependent proteinkinase from alfalfa.Plant Mol Biol 12(4), 453–61(1989).doi:10.1007/BF00017584

16

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

36. Billker, O., Lourido, S., Sibley, L.D.:Calcium-dependentsignaling and kinases inapicomplexan parasites. Cell Host Microbe5(6),612–22 (2009). doi:10.1016/j.chom.2009.05.017

37. Mihaylova, M.M., Shaw, R.J.: The ampk signallingpathwaycoordinates cell growth, autophagy andmetabolism. Nat Cell Biol13(9), 1016–23 (2011).doi:10.1038/ncb2329

38. Drewes, G., Ebneth, A., Preuss, U., Mandelkow,E.M.,Mandelkow, E.: Mark, a novel family ofprotein kinases thatphosphorylatemicrotubule-associated proteins and triggermicrotubuledisruption. Cell 89(2), 297–308 (1997)

39. Liu, F., Liang, Z., Wegiel, J., Hwang, Y.W., Iqbal,K.,Grundke-Iqbal, I., Ramakrishna, N., Gong,C.X.: Overexpression ofdyrk1a contributes toneurofibrillary degeneration in downsyndrome.FASEB J 22(9), 3224–33 (2008).doi:10.1096/fj.07-104539

40. Becker, W.: Emerging role of dyrk family proteinkinases asregulators of protein stability in cellcycle control. Cell Cycle11(18), 3389–94 (2012).doi:10.4161/cc.21404

41. Varjosalo, M., Björklund, M., Cheng, F., Syvänen,H.,Kivioja, T., Kilpinen, S., Sun, Z., Kallioniemi,O., Stunnenberg,H.G., He, W.W., Ojala, P.,Taipale, J.: Application of activeandkinase-deficient kinome collection for identificationof kinasesregulating hedgehog signaling. Cell133(3), 537–48(2008).doi:10.1016/j.cell.2008.02.047

42. Eckert, D., Andrée, N., Razanau, A.,Zock-Emmenthal, S.,Lützelberger, M., Plath, S.,Schmidt, H., Guerra-Moreno, A.,Cozzuto, L.,Ayté, J., Käufer, N.F.: Prp4 kinase grants thelicenseto splice: Control of weak splice sitesduring spliceosomeactivation. PLoS Genet 12(1),1005768 (2016).doi:10.1371/journal.pgen.1005768

43. Garrett, S., Menold, M.M., Broach, J.R.: Thesaccharomycescerevisiae yak1 gene encodes aprotein kinase that is induced byarrest early inthe cell cycle. Mol Cell Biol 11(8), 4045–52(1991)

44. van Es, S., Weening, K.E., Devreotes, P.N.: Theproteinkinase yaka regulates g-protein-linkedsignaling responses duringgrowth and developmentof dictyostelium. J Biol Chem 276(33),30761–5(2001). doi:10.1074/jbc.M103365200

45. Gui, J.F., Lane, W.S., Fu, X.D.: A serine kinaseregulatesintracellular localization of splicingfactors in the cell cycle.Nature 369(6482), 678–82(1994). doi:10.1038/369678a0

46. Giannakouros, T., Nikolakaki, E., Mylonis, I.,Georgatsou,E.: Serine-arginine protein kinases: asmall protein kinase familywith a large cellularpresence. FEBS J 278(4), 570–86(2011).doi:10.1111/j.1742-4658.2010.07987.x

47. Litchfield, D.W.: Protein kinase ck2: structure,regulationand role in cellular decisions of life anddeath. Biochem J 369(Pt1), 1–15 (2003).doi:10.1042/BJ20021469

48. Meggio, F., Pinna, L.A.: One-thousand-and-onesubstrates ofprotein kinase ck2? FASEB J 17(3),349–68 (2003).doi:10.1096/fj.02-0473rev

49. Omori, Y., Chaya, T., Katoh, K., Kajimura, N.,Sato, S.,Muraoka, K., Ueno, S., Koyasu, T.,Kondo, M., Furukawa, T.: Negativeregulation ofciliary length by ciliary male germcell-associatedkinase (mak) is required for retinalphotoreceptorsurvival. Proc Natl Acad Sci U S A 107(52),22671–6(2010). doi:10.1073/pnas.1009437108

50. Berman, S.A., Wilson, N.F., Haas, N.A., Lefebvre,P.A.: Anovel map kinase regulates flagellar lengthin chlamydomonas. CurrBiol 13(13), 1145–9(2003)

51. Jiang, Y.Y., Maier, W., Baumeister, R., Minevich,G.,Joachimiak, E., Ruan, Z., Kannan, N., Clarke,D., Frankel, J.,Gaertig, J.: The hippo pathwaymaintains the equatorial divisionplane in theciliate tetrahymena. Genetics(2017).doi:10.1534/genetics.117.200766

52. Joo, J.H., Wang, B., Frankel, E., Ge, L., Xu, L.,Iyengar,R., Li-Harms, X., Wright, C., Shaw, T.I.,Lindsten, T., Green, D.R.,Peng, J., Hendershot,L.M., Kilic, F., Sze, J.Y., Audhya, A., Kundu,M.:The noncanonical role of ulk/atg1 in er-to-golgitrafficking isessential for cellular homeostasis. MolCell 62(4), 491–506(2016).doi:10.1016/j.molcel.2016.04.020

53. Wilson, C.W., Nguyen, C.T., Chen, M.H., Yang,J.H., Gacayan,R., Huang, J., Chen, J.N., Chuang,P.T.: Fused has evolved divergentroles invertebrate hedgehog signalling and motileciliogenesis.Nature 459(7243), 98–102 (2009).doi:10.1038/nature07883

17

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

54. Russell, R.C., Tian, Y., Yuan, H., Park, H.W.,Chang, Y.Y.,Kim, J., Kim, H., Neufeld, T.P.,Dillin, A., Guan, K.L.: Ulk1induces autophagy byphosphorylating beclin-1 and activating vps34lipidkinase. Nat Cell Biol 15(7), 741–50(2013).doi:10.1038/ncb2757

55. Aranda, S., Laguna, A., de la Luna, S.: Dyrkfamily ofprotein kinases: evolutionaryrelationships, biochemical properties,andfunctional roles. FASEB J 25(2), 449–62(2011).doi:10.1096/fj.10-165837

56. Parker, L.L., Atherton-Fessler, S., Lee, M.S., Ogg,S., Falk,J.L., Swenson, K.I., Piwnica-Worms, H.:Cyclin promotes the tyrosinephosphorylation ofp34cdc2 in a wee1+ dependent manner. EMBO J10(5),1255–63 (1991)

57. Parker, L.L., Piwnica-Worms, H.: Inactivation ofthep34cdc2-cyclin b complex by the human wee1tyrosine kinase. Science257(5078), 1955–7 (1992)

58. Payne, D.M., Rossomando, A.J., Martino, P.,Erickson, A.K.,Her, J.H., Shabanowitz, J., Hunt,D.F., Weber, M.J., Sturgill, T.W.:Identification ofthe regulatory phosphorylation sitesinpp42/mitogen-activated protein kinase (mapkinase). EMBO J 10(4),885–92 (1991)

59. Roskoski, R.: Erk1/2 map kinases: structure,function, andregulation. Pharmacol Res 66(2),105–43 (2012).doi:10.1016/j.phrs.2012.04.005

60. Bian, Y., Ye, M., Wang, C., Cheng, K., Song, C.,Dong, M.,Pan, Y., Qin, H., Zou, H.: Globalscreening of ck2 kinase substratesby an integratedphosphoproteomics workflow. Sci Rep 3, 3460(2013).doi:10.1038/srep03460

61. Basnet, H., Su, X.B., Tan, Y., Meisenhelder, J.,Merkurjev,D., Ohgi, K.A., Hunter, T., Pillus, L.,Rosenfeld, M.G.: Tyrosinephosphorylation ofhistone h2a by ck2 regulatestranscriptionalelongation. Nature 516(7530), 267–71(2014).doi:10.1038/nature13736

62. Kozminski, K.G., Johnson, K.A., Forscher, P.,Rosenbaum,J.L.: A motility in the eukaryoticflagellum unrelated to flagellarbeating. Proc NatlAcad Sci U S A 90(12), 5519–23 (1993)

63. Taschner, M., Weber, K., Mourão, A., Vetter, M.,Awasthi,M., Stiegler, M., Bhogaraju, S.,Lorentzen, E.: Intraflagellartransport proteins 172,80, 57, 54, 38, and 20 form a stable

tubulin-binding ift-b2 complex. EMBO J 35(7),773–90 (2016).doi:10.15252/embj.201593164

64. Jiang, X., Hernandez, D., Hernandez, C., Ding, Z.,Nan, B.,Aufderheide, K., Qin, H.: Ift57 stabilizesthe assembledintraflagellar transport complex andmediates transport ofmotility-related flagellarcargo. J Cell Sci 130(5), 879–891(2017).doi:10.1242/jcs.199117

65. Gervais, F.G., Singaraja, R., Xanthoudakis, S.,Gutekunst,C.A., Leavitt, B.R., Metzler, M.,Hackam, A.S., Tam, J.,Vaillancourt, J.P.,Houtzager, V., Rasper, D.M., Roy, S.,Hayden,M.R., Nicholson, D.W.: Recruitment andactivation ofcaspase-8 by thehuntingtin-interacting protein hip-1 and anovelpartner hippi. Nat Cell Biol 4(2), 95–105(2002).doi:10.1038/ncb735

66. Houde, C., Dickinson, R.J., Houtzager, V.M.,Cullum, R.,Montpetit, R., Metzler, M., Simpson,E.M., Roy, S., Hayden, M.R.,Hoodless, P.A.,Nicholson, D.W.: Hippi is essential for nodeciliaassembly and sonic hedgehog signaling. Dev Biol300(2), 523–33(2006).doi:10.1016/j.ydbio.2006.09.001

67. Boudeau, J., Miranda-Saavedra, D., Barton, G.J.,Alessi,D.R.: Emerging roles of pseudokinases.Trends Cell Biol 16(9),443–52 (2006).doi:10.1016/j.tcb.2006.07.003

68. Xu, B., English, J.M., Wilsbacher, J.L., Stippec,S.,Goldsmith, E.J., Cobb, M.H.: Wnk1, a novelmammalianserine/threonine protein kinase lackingthe catalytic lysine insubdomain ii. J Biol Chem275(22), 16795–801 (2000)

69. Mukherjee, K., Sharma, M., Urlaub, H.,Bourenkov, G.P., Jahn,R., Südhof, T.C., Wahl,M.C.: Cask functions as amg2+-independentneurexin kinase. Cell 133(2), 328–39(2008).doi:10.1016/j.cell.2008.02.036

70. Shi, F., Telesco, S.E., Liu, Y., Radhakrishnan, R.,Lemmon,M.A.: Erbb3/her3 intracellular domainis competent to bind atp andcatalyzeautophosphorylation. Proc Natl Acad Sci U S A107(17),7692–7 (2010).doi:10.1073/pnas.1002753107

71. Ungureanu, D., Wu, J., Pekkala, T., Niranjan, Y.,Young, C.,Jensen, O.N., Xu, C.F., Neubert, T.A.,Skoda, R.C., Hubbard, S.R.,Silvennoinen, O.: The

18

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;

http://dx.doi.org/10.1101/168187http://creativecommons.org/licenses/by/4.0/

pseudokinase domain of jak2 is a dual-specificityprotein kinasethat negatively regulates cytokinesignaling. Nat Struct Mol Biol18(9), 971–6 (2011).doi:10.1038/nsmb.2099

72. Zeqiraj, E., Filippi, B.M., Deak, M., Alessi, D.R.,vanAalten, D.M.: Structure of thelkb1-strad-mo25 complex reveals anallostericmechanism of kinase activation. Science 326(5960),1707–11(2009). doi:10.1126/science.1178377

73. Roy, F., Laberge, G., Douziech, M.,Ferland-McCollough, D.,Therrien, M.: Ksr is ascaffold required for activation of theerk/mapkmodule. Genes Dev 16(4), 427–38(2002).doi:10.1101/gad.962902

74. Iwadate, Y., Nakaoka, Y.: Calcium regulatesindependentlyciliary beat and cell contraction inparamecium cells. Cell Calcium44(2), 169–179(2008). doi:10.1016/j.ceca.2007.11.006

75. Stanisstreet, M.: Calcium and wound healing inxenopus earlyembryos. J Embryol Exp Morphol67, 195–205 (1982)

76. McNeil, P.L., Terasaki, M.: Coping with theinevitable: howcells repair a torn surfacemembrane. Nat Cell Biol 3(5), 124–9(2001).doi:10.1038/35074652

77. Xu, S., Chisholm, A.D.: A g-alpha-q-ca2+signaling pathwaypromotes actin-mediatedepidermal wound closure in c. elegans. CurrBiol21(23), 1960–7 (2011).doi:10.1016/j.cub.2011.10.050

78. Cordeiro, J.V., Jacinto, A.: The roleoftranscription-independent damage signals in theinitiation ofepithelial wound healing. Nat Rev MolCell Biol 14(4), 249–62(2013)

79. Chambers, R., Kao, C.-Y.: The effect ofelectrolytes on thephysical state of the nerve axonof the squid and of stentor, aprotozoon.Experimental Cell Research 3(3), 564–573(1952).doi:10.1016/0014-4827(52)90080-3

19

.CC-BY 4.0 International licensepeer-reviewed) is theauthor/funder. It is made available under aThe copyright holder forthis preprint (which was not. http://dx.doi.org/10.1101/168187doi:bioRxiv preprint first posted online Jul. 25, 2017;