Accepted Article

A Novel Cryptochrome in the Diatom Phaeodactylum tricornutum Influences the Regulation of Light Harvesting Protein Levels

Matthias Juhas1, Andrea von Zadow1,3, Meike Spexard2, Matthias Schmidt1, Tilman Kottke2, Claudia Büchel1*

1

Institute of Molecular Biosciences, University of Frankfurt, Max-von-Laue-Str. 9,

60438 Frankfurt, Germany.

2

Physical and Biophysical Chemistry, Bielefeld University, Universitätsstr. 25, 33615

Bielefeld, Germany.

3

present address: Institut für Mikrobiologie und Molekularbiologie, Justus Liebig

University, Heinrich-Buff-Ring 26-32, 35392 Gießen, Germany

*Corresponding author: [email protected], Tel.: +49-69-798-29602, Fax: +49-69-798-29600 Article type

: Original Article

Running title: A Novel Cryptochrome in Diatoms

Abbreviations CPD - cyclobutane pyrimidine dimer, CPF – cryptochrome / photolyase family, DTT – dithiothreitol, FAD - flavin adenine dinucleotide , Fcp - fucoxanthin-chlorophyll proteins, Lhc - light harvesting complexes, MTHF - 5,10-methenyltetrahydrofolate This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/febs.12782 This article is protected by copyright. All rights reserved.

Accepted Article

Keywords: Algae, FAD, Lhcx, Lhcf, MTHF ABSTRACT Diatoms possess several genes for proteins of the cryptochrome / photolyase family. A typical sequence for a plant cryptochrome was not found in our analysis of the Phaeodactylum tricornutum genome, but one protein grouped with higher plant and green algal cryptochromes. This protein, CryP, binds flavin adenine dinucleotide (FAD) and 5,10-methenyltetrahydrofolate (MTHF) according to our spectroscopic studies on heterologously expressed protein. MTHF binding is a feature common to both CPD photolyases and DASH cryptochromes. In recombinant CryP, however, the FAD chromophore was present in its neutral radical state and had a red-shifted absorption maximum at 637 nm, which is more characteristic for a DASH cryptochrome than a CPD photolyase. Upon illumination with blue light, the fully reduced state of FAD was formed in the presence of reductant. Expression of CryP was silenced by antisense approaches, and the resulting cell lines showed increased levels of proteins of light harvesting complexes (Lhc), the Lhcf proteins, in vivo. In contrast, the levels of proteins active in light protection, Lhcx, were reduced. Thus, CryP cannot be directly grouped to known members of the cryptochrome / photolyase family. Of all P. tricornutum proteins it is most similar in sequence to a plant cryptochrome and is involved in the regulation of light harvesting protein expression, but displays spectroscopic features and a chromophore composition most typical of a DASH cryptochrome.

INTRODUCTION Cryptochromes and photolyases together constitute a superfamily of proteins which are found in pro- as well as eukaryotic organisms. In principle, this superfamily can

This article is protected by copyright. All rights reserved.

Accepted Article

be divided in several groups [1]. Two groups represent the photolyases, i.e. enzymes working in the repair of UV-induced DNA damage either as (6-4) photolyases or cyclobutane pyrimidine dimer (CPD) photolyases. Two further groups are cryptochrome photoreceptors found in plants and animals (type I). Those work in growth regulation and in the entrainment of circadian rhythms in plants and insects. Other animal cryptochromes (type II) act as part of the clock as transcriptional inhibitors. Moreover, sensory cryptochromes have been identified in fungi and bacteria. The last group consists of DASH cryptochromes (Cry-DASH), proteins that were shown to repair single stranded or loop-structured duplex DNA [2, 3] and are found in archaea, eubacteria, fungi, plants and some vertebrates [1, 4]. The first cryptochrome photoreceptors were found in Arabidopsis thaliana [5]. They absorb light in the blue and UV region, due to binding of flavin adenine dinucleotide (FAD) and possibly a second chromophore, 5, 10- methenyltetrahydrofolate (MTHF) to the ~500 amino acid photolyase homology region. These plant cryptochromes contain additionally a C-terminal extension, which in general exhibits a strong variation in sequence and length within the superfamily. All cryptochromes and photolyases bind FAD, albeit it is present in different oxidation states. Photolyases contain FAD in the fully reduced state in the dark [6], whereas plant cryptochromes most likely start from the oxidized state [1]. For DASH cryptochromes, mixtures of oxidation states have been found in vitro [7].

Coesel et al. [8] were the first to examine cryptochromes in diatoms, more specifically in the pennate diatom Phaeodactylum tricornutum. They described three cryptochrome / photolyase family (CPF) proteins on gene level in P. tricornutum, CPF1, CPF2 and CPF4. CPF1 was most closely related to animal cryptochromes,

This article is protected by copyright. All rights reserved.

Accepted Article

and CPF2 represented a Cry-DASH, whereas CPF4 was not closely associated to either of these groups. None of the three proteins was homologous to plant cryptochromes. Their work focused on CPF1, a very unusual cryptochrome that turned out to influence the transcription of several genes and additionally exhibited (6-4) photolyase activity. The genes influenced by CPF1 action in the light included genes of the carotenoid and tetrapyrrole biosynthesis pathways, genes of carbon or nitrogen metabolism, genes involved in the processing of genetic information, and genes for light harvesting proteins.

Light harvesting proteins in diatoms, also called fucoxanthin-chlorophyll proteins (Fcp) because of their pigmentation, fall in three groups, Lhcf, Lhcr and Lhcx. Lhcfs are the major light harvesting proteins related to the same proteins in brown algae, and Lhcr were named because of their similarity to membrane intrinsic antennae associated to rhodophyte photosystem I. Lhcx are related to LhcSR proteins of Chlamydomonas reinhardtii and work in light protection [9–13]. According to the analysis of Coesel et al. [8], all tested Lhc genes except Lhcx4 were induced by blue light. In strains overexpressing CPF1, Lhcx1-3 mRNA levels were reduced compared to wild type (WT). Also the expression of Lhcr7 decreased in these overexpressors, whereas Lhcr3 was increased. In contrast, Lhcf2 as well as Lhcx4 were not influenced by the overexpression as compared to WT.

Here we re-examined the cryptochrome sequences available in the genome of P. tricornutum, with special emphasis on plant-like sequences. The strongest homology to plant cryptochromes, although still weak with 30% identity to Arabidopsis thaliana

This article is protected by copyright. All rights reserved.

Accepted Article

cryptochrome 1 (ID AEE82696.1) in the first 500 amino acids, was found for a sequence (ID 54342) with EST support in the jgi data base (http://genome.jgipsf.org/Phatr2/Phatr2.home.html). This sequence can be retrieved under several protein IDs referring to the same gene (55577 (EST), 54342 (EST), 51782 and 53141), to which we will refer to as CryP in the following. A putative homologue, CPD2 (ID 7368), has been reported for the genome of another diatom, Thalassiosira pseudonana [14], but with a low sequence identity of only 22%. Using BLAST, sequences with considerably higher identity to CryP were found only in one diatom, Thalassiosira oceanica (EJK58052.1: 53%), and in addition in Emiliania huxleyi (Haptophytes, XP_005777116.1: 43%), Guillardia theta (Cryptomonads, XM_005834455.1: 35%), and Cyanidioschyzon merolae (Rhodophytes, XM_005536144.1: 37%). All these homologues have not yet been characterized. In P. tricornutum, a longer version of CryP is present as well, starting with an up-stream start codon and thus being around 1000 bp longer at the N-terminus (found under protein ID 45095 (EST) and 34704). According to similarity searches, the N-terminal extension has the highest similarities with a predicted methylenetetrahydrofolate reductase sequence from Ostreococcus lucimarinus. Here, we focused on the characterization of the product of the short version of CryP because it has the strongest homology of all cryptochrome/photolyase candidates from P. tricornutum to plant cryptochromes and has not been investigated before.

RESULTS Potential cryptochrome genes of P. tricornutum were identified using BLAST methods. A tree of unique sequences of representative, characterised cryptochromes and photolyases was constructed to group each of these proteins into

This article is protected by copyright. All rights reserved.

Accepted Article

one of the subfamilies (Fig. 1). Very distant members such as class II CPD photolyases [15, 16] and bacterial cryptochrome / (6-4) photolyase proteins [17–19] were omitted for clarity. The tree shows that the P. tricornutum sequence ID 54342 (CryP) is clearly distant from those of animal cryptochromes, CPF1 proteins and (64) photolyases. However, a further classification was not possible, as only some limited homology to plant cryptochromes and class III CPD photolyases was found. Thus we set out to investigate its possible role in P. tricornutum.

First we checked the expression of this gene in our P. tricornutum WT strain by PCR and detected positive signals of the expected size of 181 bp (Fig. 2A). In addition, we checked P. tricornutum cell extracts with an antibody directed against the C-terminus of CryP and could verify expression of a protein of the appropriate size for the short gene version (ID 54342) of about 67 kDa (Fig. 2B). A protein encoded by the long gene version, which includes the ~1000 bp extension similar to a predicted methylenetetrahydrofolate reductase sequence from Ostreococcus lucimarinus, has a predicted molecular weight of about 107 kDa, and was, however, not detected (Supplemental Fig. S1). After having confirmed that the gene is transcribed and the short version of the protein (CryP) is present in the diatom cells, we overexpressed 6xHis-tagged CryP in E. coli. Upon purification one single band was obtained (Fig. 2C), which again showed a strong reaction with the antibody directed against CryP (Fig. 2B).

UV/Vis absorption spectra of purified, full-length CryP were taken after different time intervals (Fig. 3). Directly after elution from the purification column, the spectrum

This article is protected by copyright. All rights reserved.

Accepted Article

showed absorption maxima at 330, 382, 590, and 637 nm. The absorption in the yellow and red spectral region is typical for a flavin neutral radical bound to a protein [20]. Additionally, the prominent peak at 382 nm is indicative of a second chromophore. After 29 d at 4°C, the absorption of the neutral radical was lost and the sample exhibited absorption maxima at 375, 446, and 470 nm. The band at around 446 nm is characteristic of an oxidized flavin. Again, the peak at 375 nm points to a second chromophore, putatively a folate.

For identification of the two chromophores, fluorescence emission spectra were recorded. To verify if FAD was bound to CryP similar to other cryptochromes and photolyases, the chromophores were isolated by heat denaturation and centrifugation. The emission spectra upon excitation at 470 nm show a maximum at 540 nm and a rise in intensity upon acidification, characteristic for FAD, but not for flavin mononucleotide or riboflavin (Fig. 4A) [21]. The folate chromophore was investigated by recording emission spectra of the holoprotein and of free MTHF in aqueous solution at pH 2 upon excitation at 380 nm and 366 nm, respectively (Fig. 4B). The strong agreement in spectral characteristics identifies the second chromophore in CryP as MTHF. In addition, the emission maximum of the holoprotein is in close agreement with that recorded of MTHF in E. coli CPD photolyase [22]. A difference is found in the presence of a shoulder at 530 nm, which is neither observed for photolyase nor for free MTHF at pH 8 [23]. The excitation spectra of free MTHF at pH 2 recorded with emission at 470 nm and 530 nm, respectively, both show a maximum at 366 nm (Fig. 4B). This demonstrates that in the CryP emission spectrum both the main emission peak around 470 nm and the shoulder at 530 nm originate from MTHF. Furthermore, a contribution by flavin to the

This article is protected by copyright. All rights reserved.

Accepted Article

CryP emission at 530 nm is unlikely, because the flavin was present in its neutral radical state which is characterized by a very weak fluorescence at ~700 nm [24]. The presence of the shoulder at 530 nm therefore implies that for MTHF aqueous solution at pH 2 more closely mimics the binding pocket of the CryP holoprotein than at pH 8.

In order to investigate the response of CryP to light in vitro, blue light illumination experiments were carried out in the absence (Fig. 5A) or presence (Fig. 5B) of the reductant dithiothreitol (DTT). The sample initially contained the FAD neutral radical and additionally to a small extent oxidized FAD as seen from the absorption at 446 nm. Without DTT, the residual oxidized FAD was completely converted by blue light into the neutral radical state. With 10 mM DTT, the photoreaction proceeded further and the absorption at 330 nm and between 550 nm and 700 nm of the neutral radical state of FAD was lost upon illumination. The final spectrum after 1200 s illumination shows the fully reduced state of FAD with maxima at ~380 and ~450 nm, similar to that of Xenopus laevis (6-4) photolyase [16] (besides some contribution by scattering). In addition, most of the absorption of the MTHF at 382 nm was lost after 1200 s of illumination as well. This photobleaching points to a photoreduction of the MTHF to 5, 10- methylenetetrahydrofolate, as it has been identified before in Arabidopsis Cry-DASH (Cry3) and E. coli DNA photolyase [25].

In order to elucidate a possible function of CryP in the regulation of gene expression, we constructed knock-down mutants in P. tricornutum using antisense and RNAi (inverted repeat) approaches (Fig. 6). The different clones generated were screened

This article is protected by copyright. All rights reserved.

Accepted Article

for reduced CryP protein expression by western blotting using the specific CryP antibody. Fig. 7A depicts exemplarily the level of CryP in one of the knock-down strains compared to WT level. After screening, three knock-down strains were selected for further experiments, NA1 (antisense construct, nitrate reductase promotor), TA3 (antisense construct, FcpA promotor) and NI8 (RNAi construct, nitrate reductase promotor). The former two were characterized by about 49-59% of remaining CryP protein, whereas NI8 showed no significant reduction in protein level and was used as a control (Fig. 7C). As already mentioned above, the product of the longer gene version expected at about 107 kDa could not be detected in WT and is thus not present or only present in minor amounts (Supplemental Fig. S1). Therefore, any reduction in the level of the longer protein cannot be demonstrated for the knock-down lines, although in principle our knock-down strategy should lead to a reduction in the level of this protein as well. It has to be emphasized that many transformants with varying expression levels were obtained for all types of knockdown approaches (data not shown). Thus the transformants used here are not representative for their promotors or knock-down approaches, but were chosen for their expression levels. Mutants were not impaired in growth compared to WT cultures and did not show any obvious microscopic features different to those of the WT cells (data not shown). In order to elucidate a possible influence of CryP on Lhc protein expression, we tested protein levels with an antibody specific for the Lhcx proteins, i. e., the proteins involved in photoprotection, and a second one specific for Lhcf1-11, i.e., the most prominent members of the Lhcf family in P. tricornutum [26]. As shown in Fig. 7B and C, the expression of Lhcf1-11 was strongly enhanced in the knock-down strains, which had reduced CryP levels as compared to WT. It has to be emphasized that this was the case in different positive mutants, thus ruling out that

This article is protected by copyright. All rights reserved.

Accepted Article

this effect was due to side effects of the transformation. In contrast to Lhcf1-11, the level of Lhcx proteins was reduced in the CryP knock-down strains, and effects were slightly smaller than for Lhcf1-11. This might be due to the fact that the contribution of Lhcx to the whole antenna protein pool is already minor in WT under the low light conditions used here for culturing.

In summary, CryP from P. tricornutum is characterized by binding MTHF as well as FAD in its neutral radical state, is photoreduced by blue light, and shows a clear influence on the protein levels of two groups of Lhc proteins.

DISCUSSION Cryptochromes and photolyases are closely related and a prediction about their function from sequence or even structure is difficult [1]. But also the strict distinction between proteins functioning as DNA repair enzymes, as light receptors, or as clock components became much less clear lately, since exceptions which act both as repair enzymes and as sensory light receptors were reported, e.g., CPF1 from the diatom P. tricornutum [8] and the later identified CPF1 from a primitive green alga, Ostreococcus tauri [27–29].

The presence of MTHF in CryP points to similarity with class I and III CPD photolyases or DASH cryptochromes, where this antenna chromophore has been extensively characterized in vitro [7, 22, 30–32]. To our knowledge, specific MTHF binding has not been reported for any other cryptochrome/photolyase family member, instead, several alternative antenna molecules were found [1]. Only plant

This article is protected by copyright. All rights reserved.

Accepted Article

cryptochromes might bind MTHF as well despite a low affinity in vitro [33, 34]. The observed complete photoreduction of MTHF is common for Cry-DASH and class I photolyases but seems to be exceptionally pronounced in CryP. A similar dose of blue light quanta at a ~20 times lower intensity but 20 times longer period did not lead to a significant bleaching in Arabidopsis Cry-DASH [25]. This observation will be further analyzed as soon as a crystal structure of CryP is available for comparison of MTHF and flavin binding sites.

The long wavelength absorption of the FAD neutral radical at 637 nm in CryP is more common to DASH cryptochromes than to CPD photolyases or plant cryptochromes. This absorption is determined by the flavin microenvironment in the protein, which can be used as an indicator for a certain protein subfamily [35]. To date, all DASH cryptochromes were reported to have their long wavelength absorption maximum at 636-640 nm similar to CryP [36–40] whereas in plant cryptochromes this is found below 600 nm [34, 41]. In contrast, in class III CPD photolyases this band maximum is found at 625 nm [31, 32] and in class I CPD photolyases at 620-625 nm [22, 42], The only exception is Anacystis nidulans class I CPD photolyase with a maximum at 634 nm [43]. Thus, the long wavelength absorbance of the CryP radical state is typical for a DASH cryptochrome, but untypical for a class I / III CPD photolyase, and not in the least similar to that of plant cryptochromes, although sequence similarity of these proteins is highest (Fig. 1). Additionally, the presence of a C-terminal extension, such as the one present in CryP (Fig. 6a), distinguishes plant Cry from class I / III CPD photolyases. We did not test for an additional DNA repair activity of CryP in vivo; this will have to be elucidated. However, the expression of CryP is not up-regulated under high light

This article is protected by copyright. All rights reserved.

Accepted Article

conditions in contrast to CPF1 of P. tricornutum [44, 45], rendering a role in UVinduced DNA repair more unlikely. This role might be taken by other candidates, which have been identified as homologues of CPD photolyases in the P. tricornutum genome, but have not been characterized yet [14].

To determine the influence of CryP on the expression levels of other proteins, especially Lhcs, knock-down lines were created. This approach in principal also reduces the mRNA levels of the longer version of the gene, but even in WT we were not able to detect the longer version on protein level. However, if it is indeed present in cells in minor amounts the effects described below could also be due to the influence of the longer version in addition to that of CryP. Our data clearly show an influence on the steady state levels of Lhc proteins. Also Cry-DASH proteins were reported to have an influence on expression levels. In Synechocystis, microarray analysis suggested that Cry-DASH acts as a transcriptional suppressor [4]. Furthermore, in dinoflagellates [46], a putative photoreceptor function was already discussed for Cry-DASH. In Arabidopsis, Cry-DASH is located in chloroplasts and mitochondria [1] and in dinoflagellates plastidic localization was detected [46]. Thus localization of CryP in vivo will have to be determined experimentally in the future. A plastidic localization would rule out direct transcriptional influence on the nuclear encoded Lhc proteins, but would still allow for retrograde signaling. Since nothing is known about the down-stream action of CryP this is highly speculative and currently no hypothesis about the signaling pathway can be put forward.

This article is protected by copyright. All rights reserved.

Accepted Article

Although we severely suppressed the amount of CryP in our transformants, any effect on cell morphology or growth was not observed despite the influence on the regulation of Lhc levels. In higher plants, Lhc protein levels do not change during the day (except for seedlings), despite obvious diurnal changes in mRNA levels [47]. The same is true for Lhcf proteins in P. tricornutum (C. Büchel, unpublished results), arguing for a severe impact of post-transcriptional events on steady state protein levels like e.g. slow turn-over rates. To elucidate the final outcome of CryP repression on the diatom cells, we measured long-term effects. Lhcf1-11 proteins are a very closely related group of antenna components. Despite their similarity, all the members of the protein family are expressed in vivo [48, 49], but occur in very different amounts, with Lhcf5 being the dominant polypeptide [50, 26]. Not much is known about factors influencing the protein levels, but it is known that the Lhcf gene family is down-regulated by high light intensity [51, 52]. This regulation affects all members to very similar extents on mRNA level [44]. On protein level Lhcf5 is reduced under high light whereas Lhcf4 is increased [26]. Thus, further work will have to clarify whether CryP influences the various members of the Lhcf family to a different extent, which would also imply a change in the composition and stoichiometry of the different Lhcf trimers found in P. tricornutum cells in vivo [26]. In contrast to Lhcf, Lhcx proteins are up-regulated in high light due to their function in photoprotection [11, 13, 44, 51–53], especially after prolonged darkness [45]. The mRNA synthesis of Lhcx1-3 was severely down-regulated in mutants overexpressing CPF1 [8]. In contrast, a reduction of these Lhcx proteins was detectable in the CryP knock-down strains.

This article is protected by copyright. All rights reserved.

Accepted Article

Overexpressing CPF1 reduced the amount of Lhcx1-3 and had no influence on the transcription of Lhcx4 as well as of Lhcf2, the only Lhcf family member tested by Coesel et al. [8]. This is opposite to the influence of CryP, where Lhcx was significantly reduced in case of diminished CryP levels and the amount of Lhcf proteins was increased. The antagonistic regulation of the same genes by two flavinbinding light receptors hampers a straight forward interpretation of physiological as well as transcript quantification data acquired after blue light exposure of diatoms. Long-term blue light adaptation was found to enhance the photoprotective state of P. tricornutum cells by increasing the Lhcx1 levels amongst other changes, which is in line with our results here. Blue light adaption thereby resembled a high light acclimation with white light [54]. P. tricornutum expresses another photoreceptor, aureochrome1a, which was shown to act as a transcription factor [55]. From the promotor sequence it was speculated that aureochrome1a might regulate Lhcx1. But in knock-down strains of aureochrome1a, the features resembling a high light acclimation were even more pronounced, pointing to aureochrome1a acting as a transcriptional repressor. It should be noted that the transcription levels of Lhcx1 were not tested in these strains [55], and that the effects of the different photoreceptors may differ between Lhcx1-4. One explanation for the antagonistic action of CryP and CPF1 mentioned above would be a different absorption spectrum of the light receptors in question, giving rise to differential regulation under changing light climates. Little is known about the absorption characteristics of CPF1 in vivo. Only two points can be addressed: When comparing the spectrum of CryP with that of recombinant CPF1 [8], it is evident that CPF1 does not contain an MTHF chromophore. MTHF might act as antenna in CryP and thereby strongly enhance the response to light at around 400 nm as compared to CPF1. Furthermore, CPF1 is

This article is protected by copyright. All rights reserved.

Accepted Article

fully oxidized after purification, whereas CryP is purified containing exclusively the neutral radical state of FAD with its absorption covering almost the full visible spectral range. Only prolonged incubation of the isolated protein in darkness leads to its oxidation. In addition, CryP responds to illumination, whereby the neutral radical state is converted into the fully reduced state in the presence of reductants. Since these were in vitro experiments, the true redox state for signaling cannot be directly inferred. One might speculate that the neutral radical state might act as dark state in CryP as it is the case with the animal-like cryptochrome (aCRY) photoreceptor from C. reinhardtii [56]. The additional absorption bands in the yellow and red range of the spectrum might make a decisive difference to oxidized CPF1 as well as to aureochrome1a [57]. We have to emphasize that this potential sensitivity of the photoreceptor at 590 nm and 637 nm cannot account for a true red light effect, which is usually probed with light of higher wavelengths (e.g. with a peak maximum at 670 nm). Such an illumination indeed had no effect on Lhcf2 transcripts [8]. The additional absorption around 550-660 nm would imply a better excitation of CryP as compared to CPF1 in vivo, since in this spectral range the chromophore is not shaded by chlorophyll and fucoxanthin, the most abundant photosynthetic carotenoid of diatoms.

In summary, CryP is another unusual cryptochrome, which groups from its sequence to plant cryptochromes (Fig. 1) and accordingly contains a ~70 amino acid Cterminal extension (Fig. 6) as opposed to CPD photolyases. Moreover, it carries chromophores and exhibits spectral features reminiscent of a DASH cryptochrome (Fig. 3 and 4) and has an influence on the expression of Lhc proteins (Fig. 7). Currently, few sequences of uncharacterized proteins having more than 30% identity

This article is protected by copyright. All rights reserved.

Accepted Article

to that of CryP have been found by protein BLAST in other diatoms, haptophytes, cryptomonads and rhodophytes. Therefore, it has to be revealed if these are CryP homologues, which might form a new subfamily of the cryptochrome / photolyase family as in the case of the CPF1 proteins.

MATERIAL AND METHODS Blast search for cryptochrome sequences in P. tricornutum and phylogenetic tree The genome of P. tricornutum v2.0 [58] was checked by tBLASTn using C. reinhardtii sequences for cryptochrome genes [59]. Phylogenetic analysis was conducted using MEGA 4 [60]. Alignment of representative protein sequences of cryptochromes and photolyases was performed using the ClustalW algorithm (Supplemental Fig. S2). The phylogenetic tree was constructed applying the neighbour-joining method [61]. A bootstrap consensus tree was generated with 1000 replications.

Blast with the sequence for the C. reinhardtii plant cryptochrome CPH1 (Chlamydomonas photolyase homologue 1), Creinhardtii|Cre06.g295200|Cre06.g295200.t1.1, gene identity 21327675 in NCBI, gave the best hit with the gene ID 55577 (estExt_Phatr1_ua_pm.C_chr_60005). Due to this similarity this sequence was chosen for further investigations. The identical gene sequence (ID 54342) was already termed CryL by Nymark et al. [44] and CPD2 by Depauw et al. [14] before, but the protein has not been characterized biochemically or functionally so far. We will refer to it as CryP from now on, because its sequence is the closest sequence to plant cryptochromes found in P. tricornutum.

This article is protected by copyright. All rights reserved.

Accepted Article

Cell cultures P. tricornutum WT (UTEX culture collection, strain 646) and knock-down strains were grown in batch cultures at 18°C under a 16 h light/8 h dark regime with 40 µmol photons m-2 s-1 white light.

RNA isolation from P. tricornutum and PCR For testing the expression of CryP in P. tricornutum, RNA was isolated using the innuPrep Plant RNA-Kit (Analytic Jena, Jena), followed by DNAse I (Fermentas) treatment and reverse transcription into cDNA with RevertAid™ Reverse Transcriptase (Fermentas) using an Oligo (dT)18 primer. The cDNA was amplified using gene specific primers for CryP (cryP-fw: CCAAAGACCGCTACTTCAACAG , cryP-rv: TCCTTTCTAATGGCCTCTACCG) and histon 4 as a positive control (primers as described in Siaut et al. [62]). The products of these PCRs were analyzed on agarose gels.

Knock-down constructs Genomic DNA was isolated with the Genomic DNA Purification Kit (Fermentas, St. Leon Rot) from WT P. tricornutum cells according to the manufacturer’s instructions. Using the following oligonucleotide primers (Seq1-fw: 5’ atatgagctcccaaagaccgctacttcaacag, Seq1-rev: 5’ atatgaattctcctttctaatggcctctaccg, Seq2-fw: 5’ cagagctcccatcaagtctcgcataacc, Seq2-rev: ccagtcgacctttctaatggcctctac) two fragments from the 3’ end including the 3’UTR of CryP were amplified, sequence1 (197 bp) and sequence 2, whereby sequence 2 is identical to sequence 1 but 66 bp shorter. Thereby SacI and EcoRI restriction sites were introduced in

This article is protected by copyright. All rights reserved.

Accepted Article

sequence 1, and SacI and SalI in sequence 2. For antisense constructs, sequence 1 was ligated into either the vector pPhaNR (gene bank accession number JN180663.1 [6], or pPhaT1 (gene bank accession number AF219942 [63]) using the SacI and EcoRI restriction sites. The resulting plasmids pPhaT1-AScry and pPhaNRAScry thus carried the antisense sequence under the control of the native FcpA or nitrate reductase promoter, respectively (Fig. 6). For stem-loop constructs, sequence 2 was ligated into the plasmid pPhaNR-AScry via the SacI and SalI restriction sites, resulting in the plasmid pPhaNR-iRcry. This plasmid carried a stem-loop construct due to the tail to tail ligation of sequence 1 and sequence 2 (Fig. 6). All constructs were allowed to amplify in E. coli under ampicillin pressure and were sequenced. Plasmid DNA was isolated and used for the transformation of WT P. tricornutum cells exactly as described by Joshi-Deo et al. [64]. Clones which could survive zeocin selection were screened for reduced expression of the CryP gene on mRNA level (data not shown).

Heterologous overexpression of CryP and isolation of the protein Out of total RNA isolated from P. tricornutum with the innuPrep Plant RNA-Kit (Analytic Jena, Jena), cDNA was synthesized. CryP was amplified using the specific primers crypC_fw; CTCTAGAATGTCGAATAGCGACCATGGTG and crypC_rev; CCTCGAGTGCGAGACTTGATGGCAGC, and the product was ligated into the pET303/CT-His vector, thus creating a C-terminal 6xHis tag. BL21 Codon+ cells were transformed with this construct and clones selected by ampicillin pressure. Successful transformation was checked by PCR. DNA was isolated from the transformants and the gene sequenced. Four bases were different from the sequence in the data base, but only two lead to an exchange in amino acids (G98

This article is protected by copyright. All rights reserved.

Accepted Article

ÆR, R568 ÆC). To check whether these changes are mistakes from amplification or due to the differences between the strain used for genome sequencing [58] and the strain used here, the procedure was repeated three times. The same changes in the sequence were found, proving that the sequence used for overexpression is correct for our P. tricornutum strain. In addition, a model of the CryP structure with the crystal structure of Arabidopsis cryptochrome1-PHR as a template (pdb 1u3d) demonstrated that R98 is predicted to be located far away from the FAD binding pocket on the surface of the protein (Supplemental Fig. S3), whereas C568 is part of the C-terminal extension. Thus, the changes are unlikely to have any impact on the function of this protein.

The transformants were cultured in DYT medium (16 g/L tryptone, 10 g/L yeast extract, 5 g/L NaCl) supplemented with 100 µg/mL ampicillin for selective pressure. At an OD600 of 0.5, the temperature was lowered from 37 to 18°C. The induction with isopropyl-β-D-1-thiogalactoside reaching a final concentration of 10 µM was carried out at OD600 of 0.8. Cells were harvested after 20 h by centrifugation (5,000 g, 20 min, 4°C). For purification, cells were disrupted using a French Press (SLM Aminco, 3 x 1,000 psi) in a 50 mM sodium phosphate buffer, pH 7.8, containing 100 mM NaCl, 20% v/v glycerol, protease inhibitor (complete EDTA-free, Roche), and DNase. For separation of cell debris the solution was centrifuged at 100,000 g and 4°C for 1 h. The supernatant was supplemented with 60 mM imidazole and applied to a His affinity column (Merck, Novagen His.Bind® Resin). Elution was carried out at 300 mM imidazole and fractions were collected. The protein was dialyzed against 50 mM phosphate buffer, pH 7.8, 100 mM NaCl, 20% v/v glycerol. For isolation and identification of the chromophores, the protein was denatured by heat (95°C, 10 min)

This article is protected by copyright. All rights reserved.

Accepted Article

and centrifuged for 2 x 10 min at 15,000 g. The supernatant was separated, and analyzed spectroscopically before and after acidification with 50 mM phosphoric acid to pH 2.1.

Spectroscopy on heterologously expressed protein UV/Vis spectra of CryP were taken at 10°C using a UV-2450 spectrometer (Shimadzu). Samples were illuminated with a blue LED at 455 nm with a full width at half-maximum of 20 nm and an intensity of ~30 mW / cm2 at the sample (Luxeon Star, Lumileds) in the absence and the presence of 10 mM DTT. Fluorescence emission spectra of protein, isolated chromophores, and MTHF (Schircks) were taken at 10°C in a JASCO FP 8300 spectrometer. MTHF was dissolved in 50 mM H3PO4, pH 1.8, to prevent degradation. All fluorescence spectra were corrected afterwards for instrumental response.

Protein isolation from P. tricornutum, SDS-PAGE and Western blot 50 ml of each cell culture was harvested by centrifugation 6 h after the onset of light after 7 days of culturing, and immediately frozen in liquid nitrogen. Total cell extracts of P. tricornutum were obtained by breaking the cells with a TissueLyzer (Qiagen) using glass beads and 700 µL lysis buffer (50 mM Tris, pH 6.8 and 2% SDS) for 10 min at 50 Hz. Insoluble material was removed by centrifugation (150 g) for 20 min at 4°C, and protein concentration in the soluble fraction determined using Roti®-Quant universal (Roth) according to the manufacturer´s guidelines.

This article is protected by copyright. All rights reserved.

Accepted Article

Proteins were separated on gels according to Schägger and von Jagow [65] and blotted onto PVDF membranes. As loading controls, gels run in parallel were Coomassie stained. Blots were decorated with either α-CryP, α-Lhcf1-11 or αFcp6(Lhcx) antibodies. Two short synthetic polypeptides specific for CryP (CTATSTVKRRKTKRTTK; TKKEYRIDRLGKVLQGC) were used together for commercial production of α-CryP in rabbits (Eurogentec, Belgium), whereby the preserum did not show any signal at the appropriate size (Supplemental Fig. S1). The antiserum of the final bleeding was used. Similarly, a sequence identical in the Lhcf1 to Lhcf11 sequences of P. tricornutum had been used for the commercial production of α-lhcf1-11 [66]. α-Fcp6 was directed against a Lhcx protein of Cyclotella meneghiniana (Fcp6), but binds Lhcx proteins of P. tricornutum as well. It was produced employing the same peptide as Westermann and Rhiel [67] (Eurogentec, Belgium). Detection was achieved using the ECL method (Amersham). For relative protein quantification, dilution series of WT and mutant proteins were compared on one blot (Fig. 7 A/B). Blots were analyzed using ImageJ and signals were normalized for small differences in loading using ImageJ on the Coomassie stained gels of the same samples. Then the values for the different dilutions from one blot (see Fig. 7A/B) were used to calculate linear regression curves for WT and knock-down line, respectively, which in turn were used to determine the ratio of proteins between WT and knock-down line for one particular blot (for details see Supplemental Fig. S4). This procedure was carried out for each protein preparation separately, whereby five preparations for the quantification of CryP and two preparations for the quantification of Lhcs were used.

This article is protected by copyright. All rights reserved.

Accepted Article

Acknowledgements: We thank Kerstin Pieper and Katharina Kuhlmeier for their excellent help regarding cloning, western blots and SDS-PAGE. TK is thankful to Thomas Hellweg for generous support. We are grateful for financial support from the Deutsche Forschungsgemeinschaft (DFG) in the framework of FOR1261 (Bu812/81,2 and Ko3580/1-2)

REFERENCES 1. Chaves I, Pokorny R, Byrdin M, Hoang N, Ritz T, Brettel K, Essen L, van der Horst GTJ, Batschauer A & Ahmad M (2011). The cryptochromes: blue light photoreceptors in plants and animals. Annu. Rev. Plant Biol. 62, 335–364. 2. Selby CP & Sancar A (2006). A cryptochrome/photolyase class of enzymes with single-stranded DNA-specific photolyase activity. Proc. Natl. Acad. Sci. USA 103, 17696–17700. 3. Pokorny R, Klar T, Hennecke U, Carell T, Batschauer A & Essen L (2008). Recognition and repair of UV lesions in loop structures of duplex DNA by DASH-type cryptochrome. Proc. Natl. Acad. Sci. USA 105, 21023–21027. 4. Brudler R, Hitomi K, Daiyasu H, Toh H, Kucho K, Ishiura M, Kanehisa M, Roberts VA, Todo T, Tainer JA & Getzoff ED (2003). Identification of a new cryptochrome class: structure, function, and evolution. Mol. Cell 11, 59–67. 5. Ahmad M & Cashmore AR (1993). HY4 gene of A. thaliana encodes a protein with characteristics of a blue-light photoreceptor. Nature 366, 162–166. 6. Sancar A (2003). Structure and function of DNA photolyase and cryptochrome blue-light photoreceptors. Chem. Rev. 103, 2203–2238. 7. Song S, Dick B, Penzkofer A, Pokorny R, Batschauer A & Essen L (2006). Absorption and fluorescence spectroscopic characterization of cryptochrome 3 from Arabidopsis thaliana. J. Photoch. Photobio. B 85, 1–16. 8. Coesel S, Mangogna M, Ishikawa T, Heijde M, Rogato A, Finazzi G, Todo T, Bowler C & Falciatore A (2009). Diatom PtCPF1 is a new cryptochrome/photolyase family member with DNA repair and transcription regulation activity. EMBO Rep. 10, 655–661. 9. Gundermann K & Büchel C (2008). The fluorescence yield of the trimeric fucoxanthin-chlorophyll-protein FCPa in the diatom Cyclotella meneghiniana is dependent on the amount of bound diatoxanthin. Photosynth. Res. 95, 229–235. 10. Gundermann K & Büchel C (2012). Factors determining the fluorescence yield of fucoxanthin-chlorophyll complexes (FCP) involved in non-photochemical quenching in diatoms. Biochim. Biophys. Acta 1817, 1044–1052. 11. Zhu S, Guo J, Maldonado MT & Green BR (2010). Effects of iron and copper deficiency on the expression of members of the light-harvesting family in the diatom Thalassiosira pseudonana (Bacillariophyceae). J. Phycol. 46, 974–981. This article is protected by copyright. All rights reserved.

Accepted Article

12. Peers G, Truong TB, Ostendorf E, Busch A, Elrad D, Grossman AR, Hippler M & Niyogi KK (2009). An ancient light-harvesting protein is critical for the regulation of algal photosynthesis. Nature 462, 518–521. 13. Bailleul B, Rogato A, De Martino A, Coesel S, Cardol P, Bowler C, Falciatore A & Finazzi G (2010). An atypical member of the light-harvesting complex stress-related protein family modulates diatom responses to light. Proc. Natl. Acad. Sci. USA 107, 18214–18219. 14. Depauw FA, Rogato A, Ribera d'Alcalá M & Falciatore A (2012). Exploring the molecular basis of responses to light in marine diatoms. J. Exp. Bot. 63, 1575–1591. 15. Yasui A, Eker AP, Yasuhira S, Yajima H, Kobayashi T, Takao M & Oikawa A (1994). A new class of DNA photolyases present in various organisms including aplacental mammals. EMBO J. 13, 6143–6151. 16. Okafuji A, Biskup T, Hitomi K, Getzoff ED, Kaiser G, Batschauer A, Bacher A, Hidema J, Teranishi M, Yamamoto K, Schleicher E & Weber S (2010). Light-induced activation of class II cyclobutane pyrimidine dimer photolyases. DNA Repair 9, 495– 505. 17. Hendrischk A, Frühwirth SW, Moldt J, Pokorny R, Metz S, Kaiser G, Jäger A, Batschauer A & Klug G (2009). A cryptochrome-like protein is involved in the regulation of photosynthesis genes in Rhodobacter sphaeroides. Mol. Microbiol. 74, 990–1003. 18. Zhang F, Scheerer P, Oberpichler I, Lamparter T & Krauss N (2013). Crystal structure of a prokaryotic (6-4) photolyase with an Fe-S cluster and a 6,7-dimethyl-8ribityllumazine antenna chromophore. Proc. Natl. Acad. Sci. USA 110, 7217–7222. 19. Geisselbrecht Y, Frühwirth S, Schroeder C, Pierik AJ, Klug G & Essen L (2012). CryB from Rhodobacter sphaeroides: a unique class of cryptochromes with new cofactors. EMBO Rep. 13, 223–229. 20. Massey V & Palmer G (1966). On the existence of spectrally distinct classes of flavoprotein semiquinones. A new method for the quantitative production of flavoprotein Semiquinones. Biochemistry 5, 3181–3189. 21. Siegel LM (1978). Quantitative determination of noncovalently bound flavins: types and methods of analysis. Method. Enzymol. 53, 419–429. 22. Jorns MS, Wang BY, Jordan SP & Chanderkar LP (1990). Chromophore function and interaction in Escherichia coli DNA photolyase: reconstitution of the apoenzyme with pterin and/or flavin derivatives. Biochemistry 29, 552–561. 23. Tyagi A, Penzkofer A, Batschauer A & Wolf E (2009). Fluorescence behaviour of 5,10-methenyltetrahydrofolate, 10-formyltetrahydrofolate, 10-formyldihydrofolate, and 10-formylfolate in aqueous solution at pH 8. Chem. Phys. 361, 75–82. 24. Kao Y, Saxena C, He T, Guo L, Wang L, Sancar A & Zhong D (2008). Ultrafast dynamics of flavins in five redox states. J. Am. Chem. Soc. 130, 13132–13139. 25. Moldt J, Pokorny R, Orth C, Linne U, Geisselbrecht Y, Marahiel MA, Essen L & Batschauer A (2009). Photoreduction of the folate cofactor in members of the photolyase family. J. Biol. Chem. 284, 21670–21683.

This article is protected by copyright. All rights reserved.

Accepted Article

26. Gundermann K, Schmidt M, Weisheit W, Mittag M & Büchel C (2013). Identification of several sub-populations in the pool of light harvesting proteins in the pennate diatom Phaeodactylum tricornutum. Biochim. Biophys. Acta 1827, 303–310. 27. Heijde M, Zabulon G, Corellou F, Ishikawa T, Brazard J, Usman A, Sanchez F, Plaza P, Martin M, Falciatore A, Todo T, Bouget F & Bowler C (2010). Characterization of two members of the cryptochrome/photolyase family from Ostreococcus tauri provides insights into the origin and evolution of cryptochromes. Plant Cell Environ. 33, 1614–1626. 28. Brazard J, Usman A, Lacombat F, Ley C, Martin MM, Plaza P, Mony L, Heijde M, Zabulon G & Bowler C (2010). Spectro−temporal characterization of the photoactivation mechanism of two new oxidized cryptochrome/photolyase photoreceptors. J. Am. Chem. Soc. 132, 4935–4945. 29. Brazard J, Ley C, Lacombat F, Plaza P, Mony L, Heijde M, Zabulon G & Bowler C (2012). Photoantenna in two cryptochrome–photolyase proteins from O. tauri: Presence, nature and ultrafast photoinduced dynamics. J. Photochem. Photobiol. A 234, 135–145. 30. Saxena C, Wang H, Kavakli IH, Sancar A & Zhong D (2005). Ultrafast dynamics of resonance energy transfer in cryptochrome. J. Am. Chem. Soc. 127, 7984–7985. 31. Öztürk N, Kao Y, Selby CP, Kavakli IH, Partch CL, Zhong D & Sancar A (2008). Purification and characterization of a type III photolyase from Caulobacter crescentus. Biochemistry 47, 10255–10261. 32. Oberpichler I, Pierik AJ, Wesslowski J, Pokorny R, Rosen R, Vugman M, Zhang F, Neubauer O, Ron EZ, Batschauer A, Lamparter T & Pastore A (2011). A photolyase-like protein from Agrobacterium tumefaciens with an iron-sulfur cluster. PLoS ONE 6, e26775. 33. Hoang N, Bouly J & Ahmad M (2008). Evidence of a light-sensing role for folate in Arabidopsis cryptochrome blue-light receptors. Mol. Plant 1, 68–74. 34. Immeln D, Schlesinger R, Heberle J & Kottke T (2007). Blue light induces radical formation and autophosphorylation in the light-sensitive domain of Chlamydomonas cryptochrome. J. Biol. Chem. 282, 21720–21728. 35. Immeln D, Pokorny R, Herman E, Moldt J, Batschauer A & Kottke T (2010). Photoreaction of plant and DASH cryptochromes probed by infrared spectroscopy: The neutral radical state of flavoproteins. J. Phys. Chem. B 114, 17155–17161. 36. Biskup T, Schleicher E, Okafuji A, Link G, Hitomi K, Getzoff ED & Weber S (2009). Direkte Detektion eines lichtinduzierten Radikalpaars in einem CryptochromBlaulichtrezeptor. Angew. Chem. 121, 411–415. 37. Sokolowsky K, Newton M, Lucero C, Wertheim B, Freedman J, Cortazar F, Czochor J, Schelvis, J. P. M. & Gindt YM (2010). Spectroscopic and thermodynamic comparisons of Escherichia coli DNA photolyase and Vibrio cholerae cryptochrome 1. J. Phys. Chem. B 114, 7121–7130. 38. Pokorny R, Klar T, Essen L & Batschauer A (2005). Crystallization and preliminary X-ray analysis of cryptochrome 3 from Arabidopsis thaliana. Acta Crystallogr. F 61, 935–938.

This article is protected by copyright. All rights reserved.

Accepted Article

39. Zikihara K, Ishikawa T, Todo T & Tokutomi S (2008). Involvement of electron transfer in the photoreaction of Zebrafish cryptochrome-DASH. Photochem. Photobiol. 84, 1016–1023. 40. Damiani MJ, Yalloway GN, Lu J, McLeod NR & O’Neill MA (2009). Kinetic stability of the flavin semiquinone in photolyase and cryptochrome-DASH. Biochemistry 48, 11399–11411. 41. Kottke T, Batschauer A, Ahmad M & Heberle J (2006). Blue-light-induced changes in Arabidopsis cryptochrome 1 probed by FTIR difference spectroscopy. Biochemistry 45, 2472–2479. 42. Ueda T, Kato A, Kuramitsu S, Terasawa H & Shimada I (2005). Identification and characterization of a second chromophore of DNA photolyase from Thermus thermophilus HB27. J. Biol. Chem. 280, 36237–36243. 43. Aubert C, Mathis P, Eker APM & Brettel K (1999). Intraprotein electron transfer between tyrosine and tryptophan in DNA photolyase from Anacystis nidulans. Proc. Natl. Acad. Sci. USA 96, 5423–5427. 44. Nymark M, Valle KC, Brembu T, Hancke K, Winge PW, Andresen K, Johnsen G & Bones AM (2009). An integrated analysis of molecular acclimation to high light in the marine diatom Phaeodactylum tricornutum. PLoS ONE 4, e7743. 45. Nymark M, Valle KC, Hancke K, Winge P, Andresen K, Johnsen G, Bones AM, Brembu T & Subramanyam R (2013). Molecular and photosynthetic responses to prolonged darkness and subsequent acclimation to re-illumination in the diatom Phaeodactylum tricornutum. PLoS ONE 8, e58722. 46. Brunelle SA, Hazard ES, Sotka EE & van Dolah FM (2007). Characterization of a dinoflagellate cryptochrome blue-light receptor with a possible role in circadian control of the cell cycle. J. Phycol. 43, 509–518. 47. Piechulla B (1999). Circadian expression of the light-harvesting complex protein genes in plants. Chronobiol. Int. 16, 115–128. 48. Lepetit B, Volke D, Gilbert M, Wilhelm C & Goss R (2010). Evidence for the existence of one antenna-associated lipid-dissolved and two protein-bound pools of diadinoxanthin cycle pigments in diatoms. Plant Physiol. 154, 1905–1920. 49. Grouneva I, Rokka A & Aro E (2011). The thylakoid membrane proteome of two marine diatoms outlines both diatom-specific and species-specific features of the photosynthetic machinery. J. Proteome Res. 10, 5338–5353. 50. Lepetit B, Volke D, Szabó M, Hoffmann R, Garab G, Wilhelm C & Goss R (2007). Spectroscopic and molecular characterization of the oligomeric antenna of the diatom Phaeodactylum tricornutum. Biochemistry 46, 9813–9822. 51. Janssen M, Bathke L, Marquardt J, Krumbein WE & Rhiel E (2001). Changes in the photosynthetic apparatus of diatoms in response to low and high light intensities. Int. Microbiol. 4, 27–33. 52. Becker F & Rhiel E (2006). Immuno-electron microscopic quantification of the fucoxanthin chlorophyll a/c binding polypeptides Fcp2, Fcp4, and Fcp6 of Cyclotella cryptica grown under low- and high-light intensities. Int. Microbiol. 9, 29–36. 53. Beer A, Gundermann K, Beckmann J & Büchel C (2006). Subunit composition and pigmentation of fucoxanthin-chlorophyll proteins in diatoms: Evidence for a

This article is protected by copyright. All rights reserved.

Accepted Article

subunit involved in diadinoxanthin and diatoxanthin binding. Biochemistry 45, 13046–13053. 54. Schellenberger Costa B, Jungandreas A, Jakob T, Weisheit W, Mittag M & Wilhelm C (2013). Blue light is essential for high light acclimation and photoprotection in the diatom Phaeodactylum tricornutum. J. Exp. Bot. 64, 483–493. 55. Schellenberger Costa B, Sachse M, Jungandreas A, Bartulos CR, Gruber A, Jakob T, Kroth PG, Wilhelm C & Campbell DA (2013). Aureochrome 1a is involved in the photoacclimation of the diatom Phaeodactylum tricornutum. PLoS ONE 8, e74451. 56. Beel B, Prager K, Spexard M, Sasso S, Weiss D, Müller N, Heinnickel M, Dewez D, Ikoma D, Grossman AR, Kottke T & Mittag M (2012). A flavin binding cryptochrome photoreceptor responds to both blue and red light in Chlamydomonas reinhardtii. Plant Cell 24, 2992–3008. 57. Herman E, Sachse M, Kroth PG & Kottke T (2013). Blue-light-induced unfolding of the Jα helix allows for the dimerization of aureochrome-LOV from the diatom Phaeodactylum tricornutum. Biochemistry 52, 3094–3101. 58. Bowler C, Allen AE, Badger JH, Grimwood J, Jabbari K, Kuo A, Maheswari U, Martens C, Maumus F, Otillar RP, Rayko E, Salamov A, Vandepoele K, Beszteri B, Gruber A, Heijde M, Katinka M, Mock T, Valentin K, Verret F, Berges JA, Brownlee C, Cadoret J, Chiovitti A, Choi CJ, Coesel S, De Martino A, Detter JC, Durkin C, Falciatore A, Fournet J, Haruta M, Huysman MJJ, Jenkins BD, Jiroutova K, Jorgensen RE, Joubert Y, Kaplan A, Kröger N, Kroth PG, La Roche J, Lindquist E, Lommer M, Martin-Jézéquel V, Lopez PJ, Lucas S, Mangogna M, McGinnis K, Medlin LK, Montsant A, Oudot-Le Secq M, Napoli C, Obornik M, Parker MS, Petit J, Porcel BM, Poulsen N, Robison M, Rychlewski L, Rynearson TA, Schmutz J, Shapiro H, Siaut M, Stanley M, Sussman MR, Taylor AR, Vardi A, von Dassow P, Vyverman W, Willis A, Wyrwicz LS, Rokhsar DS, Weissenbach J, Armbrust EV, Green BR, Van de Peer Y & Grigoriev IV (2008). The Phaeodactylum genome reveals the evolutionary history of diatom genomes. Nature 456, 239–244. 59. Mittag M, Kiaulehn S & Johnson CH (2005). The circadian clock in Chlamydomonas reinhardtii. What is it for? What is it similar to? Plant Physiol. 137, 399–409. 60. Tamura K, Dudley J, Nei M & Kumar S (2007). MEGA4: Molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24, 1596–1599. 61. Saitou N & Nei M (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406–425. 62. Siaut M, Heijde M, Mangogna M, Montsant A, Coesel S, Allen A, Manfredonia A, Falciatore A & Bowler C (2007). Molecular toolbox for studying diatom biology in Phaeodactylum tricornutum. Gene 406, 23–35. 63. Zaslavskaia LA, Lippmeier JC, Kroth PG, Grossman AR & Apt KE (2000). Transformation of the diatom Phaeodactylum tricornutum (Bacillariophyceae) with a variety of selectable marker and reporter genes. J. Phycol. 36, 379–386. 64. Joshi-Deo J, Schmidt M, Gruber A, Weisheit W, Mittag M, Kroth PG & Büchel C (2010). Characterization of a trimeric light-harvesting complex in the diatom

This article is protected by copyright. All rights reserved.

Accepted Article

Phaeodactylum tricornutum built of FcpA and FcpE proteins. J. Exp. Bot. 61, 3079– 3087. 65. Schägger H & von Jagow G (1987). Tricine-sodium dodecyl sulfatepolyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166, 368–379. 66. Juhas M & Büchel C (2012). Properties of photosystem I antenna protein complexes of the diatom Cyclotella meneghiniana. J. Exp. Bot. 63, 3673–3681. 67. Westermann M & Rhiel E (2005). Localisation of fucoxanthin chlorophyll a/cbinding polypeptides of the centric diatom Cyclotella cryptica by immuno-electron microscopy. Protoplasma 225, 217–223.

Supporting Information: Figure S1: Specificity of the antibody against CryP Figure S2: Alignment of cryptochrome and photolyase sequences for phylogenetic analysis Figure S3: 3D model of CryP and the sequence alignment used Figure S4: Quantification of proteins using antibodies directed against CryP, Lhcf111 or Lhcx(Fcp6).

This article is protected by copyright. All rights reserved.

Accepted Article

Figure 1: Phylogenetic tree of representative members of the superfamily of cryptochromes (CRY) and photolyases. CPF proteins from P. tricornutum group into the CPF1 family (CPF1), into the CRYDASH family (CPF2), and close to the CRY-DASH (CPF4). CryP investigated in this study shows some weak homology to plant cryptochromes and class III CPD photolyases. Very distant members such as class II CPD photolyases and bacterial cryptochromes / (6-4) photolyases were omitted for clarity. Bootstrap values >95% are marked by an asterisk. The scale bar represents amino acid substitutions per site.

Figure 2: CryP expression in WT P. tricornutum and over-expression in E. coli. In (a), PCR was carried out using equal amounts of cDNA from WT P. tricornutum cells with CryP and histon4 (H4) specific primers. (b) depicts immunoblots using αCryP on P. tricornutum cell extracts (WT) and the recombinant protein isolated from E. coli (R). This protein eluted as a single protein band from the affinity column as seen on SDS-PAGE (C).

This article is protected by copyright. All rights reserved.

Accepted Article

Figure 3: UV/Vis absorption spectra of heterologously expressed and purified CryP in different redox states. Directly after purification (solid line), flavin in its neutral radical state is observed as indicated by characteristic features between 550 and 700 nm. After 29 d at 4°C (dashed line), flavin is completely oxidized with an absorption maximum at 446 nm. Additionally, both spectra show the presence of a second chromophore with an absorption maximum at around 380 nm.

This article is protected by copyright. All rights reserved.

Accepted Article

Figure 4: Fluorescence spectra of the chromophores of CryP (a) The chromophores of CryP were isolated by heat denaturation and centrifugation. The supernatant was analyzed by fluorescence spectroscopy. The emission spectra with excitation at 470 nm show a maximum at 540 nm and a rise in intensity upon acidification, which are characteristic of flavin adenine dinucleotide (FAD). (b) Fluorescence emission spectrum of CryP holoprotein with excitation at 380 nm (black line). The second chromophore besides FAD is identified as 5,10methenyltetrahydrofolate (MTHF) by comparison to the scaled emission spectrum of free MTHF at pH 2.0 (blue line). Excitation spectra of free MTHF measured with emission at 470 nm (green line) and 530 nm (red line) show a maximum at 366 nm and demonstrate that in CryP both the emission peak around 470 nm and the shoulder at 530 nm originate mainly from MTHF.

This article is protected by copyright. All rights reserved.

Accepted Article

Figure 5: Blue light response of CryP Recombinant CryP was illuminated with blue light in absence (a) and presence (b) of the reductant DTT. Spectra were taken before and after blue light illumination for the indicated time intervals. Without reductant, the partially oxidized CryP is converted fully back into the neutral radical state. In presence of DTT, both MTHF and FAD are photoreduced. The final spectrum after 1200 s of illumination points to the fully reduced state of FAD overlapped by some scattering.

This article is protected by copyright. All rights reserved.

Accepted Article

Figure 6: Protein domain structure of CryP and constructs used for CryP knock-down mutants. Predicted protein domain structure of CryP (Pt54342) (a) and constructs used for transformation of P. tricornutum to yield CryP knock-down lines (b). The sequences S1 and S2 were amplified from genomic DNA (upper panel) and both were ligated tail to tail into the vector pPhat-NR (nitrate reductase promotor, PNR) for inverted repeat (IR) constructs, or only S1 was ligated into pPhat-T1 (FcpA promotor, PFcpA) or pPhat-NR to yield antisense (AS) constructs (lower panel).

This article is protected by copyright. All rights reserved.

Accepted Article

Figure 7: Protein expression in WT and CryP knock-down strains CryP, Lhcf1-11 and Lhcx expression levels in WT and the different CryP knock-down strains were evaluated by loading equal amounts of total cell protein (100 µg in case of CryP, 4 µg in case of Lhc) from knock-down lines and WT, and by dilution series to test for linearity of the detection range. Blots were then probed using α-CryP, αLhcf1-11 or α-Fcp6(Lhcx), respectively. A representative blot of the NA1 clone and WT probed with α-CryP is shown in (a), whereas in (b) α-Lhcf1-11 and α-Fcp6(Lhcx) were used. In (c) the results of CryP, Lhcf1-11, and Lhcx quantifications in NI8, TA3, and NA1 are depicted in comparison to WT. For normalization WT levels were set to 1 (indicated by the dotted line). Data are represented as mean and standard deviation of experiments in case of CryP, using five independent protein

This article is protected by copyright. All rights reserved.

Accepted Article

preparations. For statistics, values were analysed using the non-parametric Two paired samples Wilcoxon rank test, whereby * represents P < 0.05 in the comparison of NI8, NA1, and TA3 with WT, respectively. For the two independent preparations used to determine the levels of Lhcf1-11 and Lhcx, both values are depicted separately. For calculation of the ratios see Supplementary Fig. S4. .

This article is protected by copyright. All rights reserved.