Most plausibly, the prokaryotic ancestor(s) of eukaryotic chloroplasts were fully equipped with their own multifaceted signal transduction circuitry [14, 48], as observed in cyanobacterial and proteobacterial genomes which encode multiple sensor kinase/response regulator pairs [11, 12, 49]. Since the chloroplast must respond to both extra- and intra-cellular cues, one might anticipate strict conservation of these signal transduction arrays. However, given the observations that fewer sensor/response circuits exist in intracellular bacterial pathogens than in free-living representatives , one may argue that a reduction in the ancestral plastid genome size after endosymbiosis may have driven the loss of chloroplast-encoded His-to-Asp regulatory arrays (see Table 1).
The chloroplast genomes (Table 1) of several rhodophytic algae [51–56], the glaucophyte Cyanophora paradoxa , the haptophyte Emiliania huxleyi , the cryptophyte Guillardia theta , and the charophyte Chlorokybus atmophyticus  have been shown to encode the response regulator gene and in some, the sensor kinase gene for the His-to-Asp proteins. The presumptive loss of the sensor kinase seen in some chloroplast genomes may suggest that under such a circumstance, the regulatory protein may be governed by nuclear-encoded sensor kinases or by yet undescribed accessory proteins that are either of nuclear or chloroplast origin [see  for discussion]. The occurrence of this signalling array is less well documented in stramenopiles. The chloroplast genome of the raphidophyte Heterosigma akashiwo encodes a single response regulator and its cognate sensor kinase [ and this study], but the chloroplast genomes of representatives within the bacillariophytes and the pelagophytes lack both proteins of this two-component system [61, 62]. More than 75 green plant (~9 chlorophyte and ~66 charophyte) chloroplast genomes have been sequenced. Of these only the charophyte Chlorokybus atmophyticus encodes a response regulator protein. It should be noted, however, that partial footprints of (laterally transferred?) prokaryotic His-to-Asp transduction pairs have been identified in some green-plant nuclear genomes [63–66]. Such truncated His-to-Asp constructs have been shown to signal mitogen-activated protein kinase cascades, which cause the differential regulation of targeted genes [63, 67].
The taxonomic distribution of this gene-pair is poorly understood. Complete chloroplast genomes are few, especially in species-rich lineages that are represented by only a small number of complete chloroplast genomes (e.g., the bacillariophytes). Whether the maintenance of the His-to-Asp signal transduction apparatus in the chloroplast corresponds to established phylogenies remains to be determined. One might anticipate that the chloroplasts of chromophytic algae would retain this His-to-Asp array since they are the product of a serial endosymbiotic event, which involved a rhodophytic algal ancestor. Unfortunately, data for the diverse taxonomic assemblage of chromophytes has been both minimal and conflicting. The presence of the His-to-Asp array in Heterosigma akashiwo appears to reflect the retention of an ancestral signature. Whether the loss of the His-to-Asp pair in bacillariophyte chloroplast DNA represents a derived genotype, remains an open question. Regardless of phylogenetic profile, the evolutionary retention of all or part of a His-to-Asp signal transduction circuit in some distantly related algal chloroplasts strongly suggests that this biochemical mechanism must play an important role in the maintenance of chloroplast homeostasis.
We propose that the Heterosigma akashiwo Tsg1/Trg1 signal transduction pair, in concert with an RNA polymerase σ70 subunit, is involved in regulating chloroplast gene transcription. The environmental stimulus that regulates the signal transduction response remains elusive. The inability to create gene-knockout mutants or perform transformation experiments in chromophytic algae (except diatoms, which lack His/Asp genes) has hampered gene expression studies that are needed to provide direct evidence for the role of His/Asp systems in chloroplast function. However, one might infer function given the similarity of the chloroplast-encoded H. akashiwo Tsg1 and the ycf26 -encoded proteins of Emiliania huxleyi, Cyanidium caldarium, Gracilaria tenuistipitata var. liui, Porphyra purpurea and P. yezoensis to the cyanobacterial sensor kinases and NblS. Hik33 has been shown by deletion studies to impact the expression of selected genes in response to osmotic and low temperature stress [27, 68, 69], while its homologue NblS is reported to serve as a sensor of nutrient stress and high light intensity . The underlying mechanism driving these physiological responses may be governed by redox and light signals for two reasons [discussed in [27, 28]]: (a) both Hik33 and NblS possess a PAS domain, which is thought to be involved in redox and light sensing  and, (b) a large majority of the genes impacted by Hik33 and NblS are related to photosynthesis [27, 28]. It has been proposed that redox control of gene expression is a fundamental evolutionary selection mechanism responsible for the maintenance of chloroplast-encoded gene regulation systems [71, 72]. As shown in Figure 1, the H. akashiwo Tsg1 protein has a putative PAS domain. While the three dimensional structure of PAS domains is conserved among taxa, the primary protein sequences that comprise this motif are often diverse . As more types of PAS domains are characterized, the E-value for the PAS domain in H. akashiwo Tsg1 should become more robust.
An important difference between Heterosigma akashiwo Tsg1 and its homologues is the absence of a transmembrane region and a HAMP domain. The possibility of a split tsg1 gene was not supported by detailed analysis of the completely sequenced H. akashiwo chloroplast genome. The absence of a transmembrane region implies that Tsg1 is most likely present in the stroma. Though the majority of described Tsg1 proteins are putatively membrane bound (data not shown, SMART database search [23, 24]), soluble histidine kinases have been identified [74, 75]. Studies are underway using a Tsg1 peptide antibody to verify the location of the protein in the cell.
Contrary to a previous report , our data indicate that RpaB in the cyanobacterium Synechocystis shares not only similarity to the ycf27 proteins in red algae but also to H. akashiwo Trg1. The premise that ycf27 homologues are restricted to eukaryotic algae containing phycobilisomes [76, 77] is contrary to the description of this protein in the non-phycobilisome containing algae – Heterosigma akashiwo (Trg1) , Guillardia theta (ycf27) , Emiliania huxleyi (ycf27) , and Chlorokybus atmophyticus (ycf27). Nonetheless, the hypothesis that RpaB regulates the synthesis of (unknown) "factors required to couple phycobilisomes to PS1 or PSII" [76, 77] is consistent with the possible role of this protein in redox/light sensing.
The assignment of Heterosigma akashiwo Trg1 to the "Class 2" (or " ompR " super family) of transcriptional regulators offers additional insight to its function in the plastid. In prokaryotic cells, some "Class 2" proteins (such as PhoB) regulate transcription through interaction with the σ subunit of RNA polymerase  while others associate with the α subunit of this enzyme [29, 79]. We have identified a PhoB-like signature for the RNA polymerase recognition domain in Trg1 and a putative chloroplast-targeted σ70 subunit in H. akashiwo. A comparable transcriptional mechanism appears to be present in Cyanidioschyzon merolae, Cyanidium caldarium, and Guillardia theta as both PhoB-like signatures and σ subunits have been identified (Table 2) [55, 80–83].
Sigma factors in concert with core RNA polymerase selectively target chloroplast genes for transcription. For example, the prokaryotic-like, plastid-encoded polymerases with their associated σ factor(s) exclusively transcribe many genes that impact the photosynthetic process including rbcL, psbA, psbD, petB, ndhA, atpI, atpH and rps14 [84–87]. Both plastid-encoded polymerase and the phage-like nuclear-encoded polymerase can transcribe rrnA, atpB, clpP. It should be noted however, that these eubacterial RNA polymerase-associated σ factors often interact with regulatory proteins (such as Trg1) and this association may further influence the transcription of specific genes .
How could the transcription of a small, select set of genes impact chloroplast homeostasis? One might propose that a hierarchical assembly of proteins during the formation of molecular complexes could provide an exceptionally efficient mechanism for regulating the quantitative and qualitative production of molecular structures necessary for the maintenance of chloroplast function. In the chloroplast, many large functional complexes that drive oxygenic photosynthesis and carbon fixation are constructed with a definitive stoichiometry that reflects the cooperative interaction between plastid and nuclear genomes. Studies suggest that protein complex formation is regulated by the presence of a "dominant assembly partner" whose presence assures the production of its assembly associates in the correct proportions . For example, a Chlamydomonas mutant lacking the D1 protein (encoded by psbA) expresses only minimal levels of D2 (psbB) as well as CP47 (psbC) proteins. Similarly, mutants in CP47 (psbC) or D2 (psbB) show depressed concentrations of D1 (psbA) protein. In contrast, Chlamydomonas cells that were mutated in CP43 (psbD) were able to assemble D1, D2 and CP47 into a stable complex. The existence of such an assembly cascade is not restricted to photosystem II construction. When cytochrome b6 (petB) or subunit IV (petD) are not present, cytochrome f (petA) synthesis drops to 10% of that in wild type cells. A similar synthesis cascade appears also to occur in the biogenesis of ATP synthetase, which is comprised of five subunits (α,β,γ,δ,ε). Mutants lacking β will not accumulate any other core peptide. In effect, a minimal signal transduction system (encoded in the chloroplast) in conjunction with the σ subunit (encoded in the nucleus) may have given the ancestral eukaryotic cell a simple and efficient method to integrate chimeric gene sets.