Molecular Developmental Biology of Plastids

Molecular Developmental Biology of Plastids

Maize is a C4 plant.Thus，there are two distinctive developmental programs for the principal photosynthetic cells of its leaves.Mesophyll cells lack ribulose bisphosphate carboxylase but fix carbon dioxide with a different enzyme system.Bundle sheath cells fix CO2 using ribulose bisphosphate carboxylase but these cells are relatively poor in many of the components of photosystem Ⅱ.There are a number of other differences as well.

The genes for ribulose bisphophate carboxylase and the beta subunit of CF1 lie adjacent to one another on the maize chloroplast chromosome but they are transcribed divergently［45］（see Figure 1）.We have used cloned fragments from within these two genes to determine whether the levels of their transcripts are the same in bundle sheach and mesophyll cells.Mesophyll protoplasts of maize can be separated from strands of bundle sheath cells by enzymatic digestion of mesophyll cell walls，filtration，etc.RNA isolated from preparations of the two cell types was labeled and hybridized to gene sequences.Transcripts of cfl-B are abundant in plastids of both cell types but transcripts of rc-L are abundant only in bundle sheath cells.Barring strikingly different life times for these two mRNAs in the two cell types，these data indicate that the expression of these genes is regulated at transcription［46，47］.Preparations of mesophyll cells are estimated to be contaminated by about 3%of bundle sheath cells［46］.A small amount of rc-L mRNA is found in mesophyll cell preparations；this may be contamination or there may be a small amount of transcription of rc-L in mesophyll cells.The mechanisms for control of expression of these two genes in mesophyll vs bundle sheath cells are obviously of considerably interest.

Photoregulated maturation of plastids is another interesting developmental program.In general，angiosperm seedlings grown in darkness on seed reserves are pale yellow.These etiolated plants contain plastids（etioplasts）that are arrested at a characteristic stage of development.Etioplasts lack chlorphyll and many of the proteins and lipids required for photosynthetic function.Upon illumination，or soon afterward，chlorophyll accumulates together with new proteins and lipids；low levels of photosynthetic competence can be detected within a few hours after illumination of seedlings although complete maturation of the chloroplast takes considerably more time.

Chloroplast DNA has been hybridized with RNA from etioplasts or RNA from chloroplasts to determine what fraction of the chromosome is transcribed in illuminated plants but not dark-grown seedlings.About 14%of the maize plastid chromosome that is transcribed in chloroplasts was found to be untranscribed in etioplasts［L.Haff and L.Bogorad，unpublished］.In these and some other experiments，RNA pools are being measured and thus it is more strictly correct to state that etioplast RNA lacks transcripts of about 14%of the chromosome that are present in the chloroplast RNA population.Considering other limitations of the experiments，it is more appropriate to state that transcripts of about 14%of the chromosome are either absent or are present in much lower amounts in etioplast than in chloroplast RNA.

The first gene whose transcript level was found to be raised upon illumination of seedlings，ie，a“photogene”was identified by a combination of techniques.These included comparison of the hybridization of radioactive RNAs from etioplasts and chloroplasts against fragments of chloroplast DNA generated by a restriction endonuclease and translation of these RNAs in vitro to identify differences in their polypeptide products［32］.These approaches were combined with analyses of proteins made by isolated chloroplasts and by intact plants at various stages during light induced development［48，49］.From all of these studies it was concluded that the primary translation product of this maize photogene is a 34.5 kDa（kilodalton）thylakoid membrane protein that is processed to 32 kDa after association with the membrane.The gene was consequently designated photogene 32.Isolated maize plastids can synthesize the 34.5-kDa primary translation product（or，at least，can complete the synthesis of started polypeptidc chains）but have little capacity for processing it to its mature 32-kDa form［49］.Photogene 32 was mapped to a portion of Barm fragment 8 of the maize plastid chromosome.When intact green leaves are supplied with 35 Smethionine，the first membrane protein to be labeled is the product of photogene 32.Since there seems to be a relatively small amount of this protein in membranes，despite its rapid labelling，it seemed reasonable to conclude that the protein must turn over relatively rapidly.The photogene 32 product has been shown subsequently to bind azido-atrazine［51］and has been judged to function in electron transport at the QB site in the photosynthetic electron transport chain.

We have recently gone on to map additional maize plastid photogenes—ie，genes for which more transcripts are present during light-induced development than in darkness-by improved methods for detecting them by hybridization against chloroplast DNA fragments.Furthermore，we have used cloned fragments of the maize chloroplast chromosome rather than fragments of total plastid DNA as sources of DNA for localizing of photogenes on the map of the chloroplast chromosome.The latter has two advantages：① confusion is avoidedtwo or more sequences of a single size class can lie in a single band after agarose gel electrophoresis of total plastid DNA digested with a restriction enzyme，but the likelihood of this happening is sharply reduced when smaller portions of the chromosome are used；②a single DNA fragment can be further analyzed in detail in hybridization experiments.Three well-separated regions of the maize chloroplast chromosome have been shown to harbor“highly responsive”photogenes.In addition，a“minor photogene”regions，sequences for which transcript levet changes are smaller than for the“major photogene regions”has been found at a fourth location in the small unique DNA region of the chromosome（see Figure 1）.Thus，there is not a single photogene“operon”in the maize chloroplast chromosome but there are several regions from which transcription is stimulated by illumination of darkgrown seedlings.These photoresponsive regions comprise about 19%of the unique DNA（counting only one inverted repeat）of the maize chloroplast chromosome.This number is amazingly close to the roughly 14%estimated at from saturation hybridization data.

One objective of mapping photogenes is to open the way to studying the mechanisms of photoregulation of transcription.Now that we know that there is not a single photogene operon，we must ask whether all of the plastid photogenes of maize are of a single expression class despite this.Is transcription of all photogenes stimulated to the same extent？Is the interval between initiation of illumination and increase in the transcript pool size the same for all photogene regions？Is transcription of all photogenes affected in the same manner when plants thar have been illuminated are returned to darkness？

These questions were addressed by using cloned fragments of photogenes to measure the amounts of complementary RNA present in seedlings at different stages of light-induced development using either the dot blot［51］or Northern hybridization methods.We learned from the dot blot hybridization experiments that there are differences in the kinetics of induction of transcription of the various photoregulated regions.We have also found that under our conditions transcript pools of all photogenes reach maxima at between 10 and 44 h after the commencement of illumination and then，eicher in light or darkness，the size of each pool drops.The singular exception to this is the level of transcripts of photogene 32.

To study these problems，RNA was isolated from dark-grown maize seedlings and from seedlings of the same growth lot illuminated for 2，10，20，40 or 68 h：some samples were returned to darkness after 20 h of illumination.In other experiments，harvests were made after dark-grown seedlings had been illuminated for 2，4，6，or 10 h or for 3，6，12，18，24，or 36 h.Aliquots of RNA samples were affixed to Gene Screen filters（New England Nuclear.Waltham.Massachusetts）.Appropriate cloned DNA fragments were nick-translated to provide 32P-labeled probes to measure the abundance of specific RNA species in each sample using the dot blot hybridization procedure.In other experiments，the Northern procedure was used；ie，RNA was denatured and subjected to agarose gel electrophoresis，the RNAs were transferred to Gene Screen，and nick-translated probes were used to identify bands of complementary RNA.In the latter case，the sizes of transcripts could also be established.(https://www.daowen.com)

Careful and sensitive analyses demonstrated that transcripts of photogenes were present in etiolated plants but that these increased in abundance during illumination.For example，RNA complementary to Bam fragments 17 and 21 increased about four fold in abundance during the first 10 h of illumination.RNA from Bam fragments 3 and 24 increased to about the same extent but only after 20 h of continuous illumination.One striking observation is that transcripts of three of the four photogene regions studied in detail decrease in abundance starting after ten or twenty h of continuous illumination of the dark-grown seedlings.By 44 or 68 h of illumination the RNA transcribed from these regions was only about as abundant as the same species in etiolated plants.The steady-state level of transcripts for rc-L，for example，was slightly below that in etiolated plants after the seedlings had been illuminated continuously for 44 h.An exception to this pattern is the transcript pool of photogene 32.When seedlings were returned to darkness for 20 h and maintained there for another 48 h，the pool sizes for photogene 32 mRNA remained about constant；the sizes of these pools continued to increase until 68 h（the termination of the experiment）when seedlings were maintained in the light.

We conclude that not all photogenes are the same expression class.They differ with regard to ① whether mRNA pool sizes increase throughout the first 68 h of illumination，as is the case for photogene 32，or ② whether a peak is reached at 10 or 20 h after illumination and pool sizes decline after that.The latter class can be further subdivided depending upon the time at which the maximum size of the pool is reached under our illumination conditions.

Another striking set of observacions is that，except for cfl-BE，for which only a single size of transcript is seen in maize，each photogene region is represented by from two to 13 transcripts of different size classes.

The simplest case of multiple-sized transcripts is of the gene for the large subunit of RUBISCO.The most abundant transcript of rc-L in etiolated plants is 1.6 kb，but a trace of a 1.8 kb RNA is visible in Northern hybridizations.After only 2 h of illumination，the 1.8-kb transcript is prominent and after 20 h it comprises about one-third of this RNA complemencary to re-L.But by 44 h the level of the 1.8-kb transcript is down to perhaps 20%of its maximum level and 24 h later its contribution to the total is very small（The“1.8-kb”transcript is longer at its 5'end than the 1.6-kb transcript.）.

More complex patterns of RNAs are seen for the three photogene regions we have studied that yield polycistronic mRNAs.For each of these regions there is a transcript of about 5 kb that includes messages for at least two different genes.In addition，these regions，or parts of them，are represented by a larger number of transcripts of different sizes.In one case，13 different transcript size classes have been identified as complmentary to various regions across a stretch of the plastid chromosome of about 5—6 kbp.These transcripts，like those of rc-L，are not present equimolar amounts at each stage of light-induced development.Some of the smaller transcripts differ from the largest one representing that region at the 5'end and other less than maximum-sized transcripts are shorter at their 3'ends.We do not know whether the 5'differences result from the use of multiple transcription initiacion sites，purposeful processing，or degradation.Neither do we know whether the 3'differences can be attributed to two or more possible different termination sites，purposeful processing，or degradation.Differences in abundance of some of the multiple transcripts from regions that contain more than one gene can be imagined to reflect a mechanism for controlling the relative abundance of translatable messages for different genes within a single operon.Whether this is true remains to be determined.

One photogene operon contains coding sequences for cfl-A and cfo-Ⅲ（We suspect that there are other genes in this region that are transcribed into the same 5 kb RNA but the sequencing has not been completed.）.Another maize photogene region contains fused coding sequences for the beta and epsilon subunits of coupling factor CF1［34］.Changes in RNA pool size in response to illumination and development，including the decreases seen after 20 h in the light，are parallel for the two DNA sequences that code for polypeptide components of the CF0-CF1complex，ie，cfl-BE and cfl-A plus cfo-lll.The conserved sequence resembling a prokaryotic promorer 5'to the transcribed portion of cfl-BE is known but that for the cfl-A plus cf0-1ll is not yet known.It will be interesting to see if there are common sequences that serve as part of the expression controlling mechanisms and，in general，to understand how expression of these two operons is coordinated.

The observations on photogene transcription in vivo outlined above lead to some questions.What are the mechanisms for bringing about increases in the sizes of pools of photogene transcripts？What determines the maximum sizes and the time at which a pool reaches that size？What signal or system leads to a reduction in pool size？What determines the steady-state pool size in mature plastids？Some aspects of these questions are now reasonably well-defined in outline：dark-grown maize seedlings have low levels of transcripts of every light-inducible gene we have studied；increases in the sizes of transcript pools for all of the photogenes can be detected within about 2 h after illumination-there may be a single underlying mechanism（e.g.，a change in RNA polymerase activity）for increasing transcription of all photogenes or there may be a number of signals all of which appear at the same time，assuming that the rate of RNA turnover is unaltered；photoregulated transcription may involve alternate promoters or terminators or changes in RNA processing or degradation；photogenes and photogene operons differ from one another with regard to the time at which pool sizes reach a maximum and whether they subsequently remain constant or decrease in darkness or in the light-decreases in sizes of the pools may be under negative or positive control.Repression can be imagined via a feedback loop as might be found already in etioplasts lacking products of the nuclear-cytoplasmic system and again after the gene products have accumulated（in mature plastids）or，as a positive（i.e.，derepression or enhancement）control model.The level of accumulated transcripts might be related to sizes of pools of nuclear gene specific products，eg，components of mnultimeric systems in the plastid.We have no idea of how the nuclear-cytoplasmic system might sense the completion of plastid components.Nor do we know whether the cytoplasmic products are produced all the time or are turned over after plastids mature.

Other related problems raised by the observations described here have been mentioned earlier.How are the difterent sizes of transcripts produced from a single transcription unittranscription start or stop site differences，processing，degradation？Is there any functional significance underlying this phenomenon？Some answers will come from analyses of the transcripts themselves.Other answers will come from research on transcription and processing.