kleptoplasts

ARE THE KLEPTOPLASTS FUNCTIONAL? 
We have evaluated the extent and longevity of symbiont activity by analyzing or measuring the following: microscopy images (electron, fluorescence, phase-contrast); oxygen exchange, 14CO2 fixation, photosynthetic electron transport; enzyme activities; constitutive and de novo chloroplast polypeptide banding patterns in Elysia and Vaucheria revealed by protein staining, autoradiography and immuno-blotting; accumulation of chloroplast transcripts; and chloroplast protein import.  The results of all of these experiments reveal a significant number of processes related to gene expression, protein biosynthesis, light energy capture/transfer and CO2 fixation/reduction are all occurring in the symbiotic plastids for several months, enabling long-term photosynthetic activity and survival and reproduction of the animal.

WHAT SUSTAINS THE LONG-TERM SYMBIONT ACTIVITY?
Considering the overwhelming dependency of plastid function on nuclear encoded proteins, the level of chloroplast activity observed in the animal cell is quite remarkable.  Possible factors contributing to this long-term functional association include the following:

1.  Algal nuclei in the sea slug?
We have ruled out the presence of an algal nucleus/nuclear genome in the sea slug based on electron and fluorescence microscopy observations and negative results from probing sea slug DNA with a V. litorea nuclear DNA gene probe to the multi-copy ribosomal intertranscribed spacer (ITS) region. PCR analysis was also carried out using primers designed against the V. litorea ITS and again negative results were obtained for animal and egg DNA supporting the absence of V. litorea nuclei in the sea slug.

2.  Autonomous chloroplasts?
Originally, the possibility was considered that the chloroplast genome of V. litorea possessed an unusual coding capacity enabling it to code for proteins that are typically encoded by the nuclear genome in other photosynthetic organisms.  To address this, the 115 kb chloroplast genome of V. litorea was sequenced and found to contain only 137 protein coding genes and all of the plastid rRNA genes.  In terms of gene content, V. litorea cpDNA was found to be intermediate to Porphyra, a red alga and Odontella, a diatom.  Some of the photosynthetic genes present in the cpDNA of V. litorea but not in land plant chloroplast genomes include: rubisco small subunit (rbcS), rubisco expression protein (cfxQ), several psa and psb genes, chaperonins DnaK (dnaK) and GroEL (groEL), ATPsynthase CF1 subunits (atpD and atpG) and protein elongation factors Tu (tufA) and Ts (tsf).  However, the V. litorea chloroplast genome is very much a “normal” non-green algal plastid genome and does not have any extensive or particularly unusual coding capacity to substitute for all or even a significant number of essential nuclear-encoded chloroplast-targeted proteins necessary to sustain the observed chloroplast activity.

3.  Unusual chloroplast stability
Upon isolation, V. litorea plastids display an unusual stability compared to plant plastids.  Two days after isolation, intactness remained at 90% of the original value, as measured by phase contrast microscopy.  After 3 days, about 55% of the plastids were still intact.  Functionally, the isolated plastids exhibited high rates of 14CO2 fixation; 50 and 25% of the control value was maintained after 2 and 3 days, respectively.  In vitro translation activity also remained high in isolated plastids and no differences were observed in thylakoid or stromal translation product patterns even after 3 days.  Refer to Green et al. paper, 2005

4. Unusual protein/transcript stability
Evidence to date from Western blotting indicates that several plastid encoded proteins involved in photosynthetic electron transport (PSI and PSII complexes) and CO2 fixation (rubisco subunits) are present for several months in E. chlorotica (Green et al., 2000).  In addition, PRK, light harvesting proteins (Lhcp I and II), and a photosynthetic electron transport component (PetC), all nuclear-encoded chloroplast proteins, are also present and functional after several months starvation of the sea slugs (Pierce et al., 1999, 2003; Rumpho et al., 2001).  If the genes for these essential chloroplast proteins that are unique to photosynthetic organisms are not present in the nuclear genome of the sea slug by way of horizontal gene transfer or substitution, then the possibility of unusual stability of proteins must be considered.

Chloroplast protein stability lasting for nine months as proposed for sea slug kleptoplasts is unprecedented in large part due to the constant exposure of chloroplasts to light damage or photoinhibition.  Although the sea slugs have some ability to shade themselves from excessive light in the sea by burrowing into the sediments, hiding under the algal mats, and reducing body exposure to the sunlight by closing their parapodia, they cannot fully escape the light energy in the illuminated aquaria.  Protein turnover is essential for recovery from photoinhibition, in the removal of inactive or improperly assembled proteins, as well as for the maintenance of stoichiometry of multiple-subunit complexes (some of which may be a mix of nuclear- and plastid-encoded subunits) during times of limited availability of co-factors, acclimation to environmental stress, and recovery from heat denaturation.  Both protein degradation and synthesis are indispensable to prevent the aggregation of damaged proteins and to allow for the re-synthesis/reassembly of undamaged proteins.  Turnover rates have been analyzed for only a few proteins, primarily those most sensitive to photodamage such as D1.  Protein degradation or turnover rates have not been specifically examined in chromophytes, but sequencing of the V. litorea chloroplast genome has revealed the presence of one open reading frame with homology to the nuclear encoded protease FtsH and one ORF with homology to the Clp protease regulatory subunit ClpC, both nuclear-encoded in land plants.  FtsH primarily degrades inactivated D1 which can be replaced by de novo synthesis in the kleptoplast.  The Clp protease complex (requiring both ClpP and ClpC subunits) has been implicated in the degradation of abnormally folded, unassembled, and/or inactive proteins.  In land plants ClpC is nuclear encoded and ClpP is plastid encoded; opposite of what is observed for V. litorea.  The absence of a gene source for the proteolytic subunit ClpP in the sea slugs may inhibit the formation of a functional enzyme complex in the kleptoplasts.  No other proteases are present in the V. litorea cpDNA, suggesting that those proteases involved in acclimation-degradation (nuclear-encoded DegP and an unidentified protease that degrades light harvesting proteins under high light conditions) in other organisms may not be functional in the kleptoplasts.  It remains to be determined if a low level of protein turnover is typical of V. litorea chloroplasts in general or if it is an adaptation of the kleptoplasts to life in an animal cell, in part as a result of the loss of functional chloroplast protease complexes and better absorption of reactive oxygen species.

5.  Redirection of animal proteins
The possibility that some animal nuclear encoded gene products are targeted to the chloroplast is supported by two observations: 1) sea slugs labeled with  35S-methionine in the presence of chloroplast gene expression inhibitors followed by isolation of the plastids revealed several proteins synthesized in the cytosol and recovered from the plastids and 2) radiolabeled translation products of either Elysia or Vaucheria RNA incubated with isolated V. litorea plastids resulted in a single 45 kD polypeptide imported in both cases.  Proteins destined for the mitochondria or involved in animal gluconeogenesis and pentose phosphate pathways may be directed to the foreign plastid.  Candidates include the Rieske Fe/S protein which is structurally and functionally related to the nuclear encoded mitochondrial ubiquinone-oxidoreductase complex Rieske protein, the regulatory γ -subunit of the ATP synthase complex and enzymes of the photosynthetic carbon reduction (PCR) cycle.

6.  Horizontal gene transfer
Probably the most intriguing possibility is that horizontal gene transfer (HGT) from Vaucheria to the mollusc has occurred and these nuclear encoded genes are being retargeted to the endosymbiont.  The only enzyme in the PCR cycle unaccounted for in either the plastid genome or animal genomes is phosphoribulokinase (PRK).  Since E. chlorotica has demonstrated the ability to carry out CO2 fixation for several months and PRK is essential for regenerating the CO2 acceptor ribulose-1,5-bisP (RuBP) to keep the PCR cycle turning, the gene for PRK makes an excellent candidate for HGT from the algal nucleus to the sea slug.  We have cloned the PRK gene from V. litorea and found two partial PRK gene fragments in DNA isolated from sea slugs and sea slug eggs. We are also pursuing other nuclear encoded proteins essential for plastid function, including psbO.

Eukaryotic cells and genomes have evolved from more than one prokaryote and are thus chimeras.  Increasing evidence indicates that HGT (the exchange of genes between distantly related species) has played a key role in evolution.  Gene transfers from organelle genomes to the host nucleus have contributed much to the present day eukaryotic genome.  For example, in the genome of Arabidopsis thaliana, approximately 4500 genes (18% of the genome) are of cyanobacterial origin.  In organisms that acquired plastids by secondary endosymbiosis, genes have been transferred from both the nucleus and chloroplast of the secondary symbionts to the new host.  Proposed pathways for gene transfer include the ‘bulk-transfer’ of large fragments of organellar DNA or RNA intermediates that escaped membrane constraints and recombined with the nuclear genome.  In plants, the cytochrome oxidase subunit II (COXII) gene was transferred from the mitochondrial genome to the nuclear genome via an RNA intermediate (Nugent and Palmer, 1991).  In some instances, multiple independent transfers of the same gene have occurred, e.g., the information to encode the ribosomal protein RPS10was transferred from the mitochondria to the nucleusnumerous independent times via an RNA intermediate (Adams et al., 2000).  Examples of both bulk transfer of DNA and fragments of DNA have also been documented in various organisms including Arabidopsis in which the entire mitochondrial genome has been integrated into chromosome 2 of the nuclear genome (Lin et al., 1999).

A similar process of bulk HGT from the alga to the sea slug can be imagined in kleptoplastic associations.  During the process of feeding on the algal filaments, it is very likely that nuclei passed through the digestive system of the sea slug and were broken open.  This would have presumably enhanced the transfer of algal nuclear genes or transcripts, including those encoding chloroplast-targeted proteins, to the nucleus of the sea slug.

In summary: the acquisition, incorporation and retention of intact algal plastids by E. chlorotica may be aided by the robustness of the V. litorea plastids.  However, long-term functional activity of the plastids appears to be supported by both protein stability and contributions from the sea slug itself.  It is apparent that the symbionts both fuel the sea slugs and protect them, thus driving the formation of this association.  But, of what advantage is the association to the alga?  Does feeding on algal filaments encourage fragmentation and increased propagation of the alga in nature? Whether this association will evolve to become a permanent association, remains to be seen.