**4. Structure and function of CP43**

56 Artificial Photosynthesis

partially restored by high concentrations of calcium and chloride (Popelkova & Yocum, 2007). The absence of these proteins also makes the Mn4CaO5 cluster vulnerable to the

The amino acid residues in the first and second coordination sphere are labeled. Please note that D342:D1 is obscured in this view by the metal cluster. The manganese ions are shown in purple and are labeled. Calcium is shown in cyan and the oxygens are shown in red. The

In cyanobacteria, deletion of the PsbU protein results in a mutant capable of growth in normal media, but which shows a reduced growth rate in media depleted of chloride (Shen et al., 1997). This mutant has reduced rates of oxygen evolution, a more stable S2 state, and is more susceptible to damage to its PSII centers by both light and heat stress (Shen et al., 1997; Shen et al., 1998; Inoue-Kashino et al., 2005). Mutants constructed by deletion of the extrinsic cytochrome c550 protein, encoded by the *psb*V gene, do not grow well photoautotrophically, and exhibit no capacity for photoautotrophic growth in media depleted of either calcium or chloride (Shen et al., 1998; Katoh et al., 2001). The *psb*V deletion mutants also show decreased ability to evolve oxygen and reduced numbers of PSII centers in the thylakoids (Shen et al., 1995; Katoh et al., 2001). They also exhibit S-state cycling defects, including an increase in miss and double hit factors and a slowing of the S3 → S4→ S0 transition (Shen et

These observations are all supportive of a role of the extrinsic proteins in maintaining the proper ionic environment around the Mn4CaO5 cluster. How chloride is actually bound and transported to the active site is currently unknown. Interestingly, the current crystal structure identified a third chloride ion bound near the C-terminal amino acid of PsbU, which is located between PsbU and cytochrome c550 (Umena et al., 2011; Kawakami et al., 2011). This chloride is ligated by water molecules and lies close to the exit of a proposed hydrogen-bonding network leading from the Mn4CaO5 cluster to the lumen that could possibly serve to transport anions, water or protons (Gabdulkhakov et al., 2009; Vassiliev et

This figure was generated in Pymol from the crystal structure of *Thermosynechococcus* 

2, respectively.

effects of exogenous reductants (Ghanotakis et al., 1984).

Fig. 1. Coordination Environment of the Mn4CaO5 Cluster.

*vulcanus* (PDB: 3ARC; Umena et al., 2011).

al., 1998).

al., 2010; Kawakami et al., 2011).

proximal and distal chlorides are shown in yellow and labeled Cl-1 and Cl-

The product of the *psb*C gene, the CP43 mature protein from *Synechocystis* 6803 consists of 460 amino acid residues and is well conserved. The nucleotide sequence of *psb*C has been determined from more than 50 species. The predicted plant apoproteins show approximately 95% homology with each other, and 85% and 77% homology with *Chlamydomonas* and cyanobacterial CP43, respectively. Early crystallographic studies confirmed that CP43 contains six transmembrane alpha helices (Zouni et al., 2001; Kamiya and Shen, 2003; Ferreira et al., 2004). The transmembrane helices of CP43 contain a number of conserved histidyl residues that function as chlorophyll-*a* ligands. Replacement of these histidyl residues with either tyrosyl or arginyl residues (Shen et al., 1993) results in mutants showing decreased PSII stability and defects in light-harvesting efficiency. Thirteen chlorophyll-*a* molecules associated with CP43 were assigned in the most current crystal structure (Umena et. al., 2011).

In addition to the six transmembrane alpha helices, CP43 contains five hydrophilic loops that connect the membrane-spanning domains. The positions of these loops are as follows (*Synechocystis* 6803 numbering): Loop A, Glu58-Phe98; Loop B, Gly123-Thr145; Loop C, Gly170-Asp219; Loop D, Lys242-Tyr258 and Loop E, Asn280-Arg410. Loop E, with 130 amino acids, is quite large and contains approximately 30% of the amino acids in this protein. Loops A, C and the large extrinsic Loop E are exposed to the lumenal side of the thylakoid membrane. The N- (Val1-Lys36) and C-termini (Arg448-Asp460), in addition to Loops B and D, face the stromal surface of the membrane (von Heijne and Gavel, 1988; Sayre and Wrobelboerner, 1994;). The CP47 protein has a similar structure and also possesses a large extrinsic loop.

All crystal structures of dimeric PSII show that CP43 is positioned on the D1 side of the D1- D2 heterodimer and CP47 on the D2 side. The two PSII monomers are related by a local-C2 rotation axis oriented perpendicular to the plane of the membrane. CP47 lies at the interface of the two monomers while CP43 is located at the periphery. This location is speculated to be necessary for the removal and replacement of damaged D1 during photoinhibition. In both CP43 and CP47, the six transmembrane helices are arranged as a "trimer of dimers". The large extrinsic loop of CP47 lies adjacent to the PsbO protein, consistent with a large body of biochemical evidence (Bricker & Frankel, 2002), and is also close to the extrinsic PsbU protein. The large extrinsic loop of CP43 is located adjacent to cytochrome c550 and PsbU, as well as the manganese cluster.

Early mutagenesis and biochemical studies clearly outlined a role for CP43 in oxygen evolution and the stable assembly of the PSII complex. CP43 is required for optimal oxygenevolving activity from isolated PSII preparations (Bricker, 1990). *Chlamydomonas* strains incorporating mutations that affected either the synthesis or stability of CP43 were deficient in PSII activity (Rochaix et al., 1989). In these mutants, levels of the other PSII core proteins were severely reduced. *Synechocystis* mutants lacking the *psb*C gene as a result of deletion mutagenesis accumulated PSII core complexes (minus CP43) to only 10% of wild-type levels and could not grow photoautotrophically or evolve oxygen (Rogner et al., 1991; Carpenter et al., 1990). Interruption of the *psb*C gene by insertional mutagenesis also produced a mutant incapable of evolving oxygen, but which exhibited primary charge separation (Vermaas et al., 1988). A S*ynechocystis* mutant recovered by nitrosoguanidine mutagenesis was unable to evolve oxygen or support PSII electron transport from water to either dichlorobenzoquinone or methyl viologen (Dzelzkalns & Bogorad, 1988). Further characterization of this mutant

Mutations in the CP43 Protein of Photosystem II

Affect PSII Function and Cytochrome C550 Binding 59

control rates at saturating light intensities. However, they show an enhanced susceptibility to photoinactivation. Additionally, the fluorescence rise time for this mutant was increased by a factor of two over the control. Thus, the PSII centers in this mutant do not function normally, even under normal growth conditions. Significantly, when the R320S mutant was grown in media depleted of chloride (30 M chloride), it exhibited a severely reduced photoautotrophic growth rate. The effect of chloride depletion on the growth rate of the mutant was reversed by the addition of 480 M bromide to the chloride-depleted BG-11 media. It is well known that bromide will functionally replace chloride at the oxygenevolving site in cyanobacteria grown on bromide-containing media (Yachandra et al., 1993). Additionally, it is well known that bromide can replace chloride in supporting oxygen evolution in isolated thylakoids (Kelly & Izawa, 1978) and PSII membranes (Sandusky & Yocum, 1983). Oxygen evolution rates for the mutant were further depressed to about 22% of the rate observed in control cells under chloride-limiting conditions. Addition of bromide restored these rates to those observed under chloride-sufficient conditions. The mutant exhibited a significantly lower relative quantum yield for oxygen evolution than did the control strain and this was exacerbated under chloride-limiting conditions. Fluorescence yield measurements indicated that both the mutant and control strains assembled fewer PSII reaction centers under chloride-limiting conditions. The reaction centers assembled by the mutant exhibited an enhanced sensitivity to photoinactivation under chloride-limiting conditions, with a t1/2 of photoinactivation of 2.6 min under chloride-limiting conditions compared to a t1/2 of 4.7 min under normal growth conditions. The mutant also exhibited an enhanced stability of its S2 state and an increased number of centers in the S1 state following dark incubation. These results indicate that the mutant R320S exhibited a defect in its ability to utilize chloride in support of efficient oxygen evolution in PSII. This was the first mutant of this type described in the CP43 protein**.** In terms of this chloride effect, R320S has a similar phenotype to the R448G, K321G, and F363R mutants constructed in the CP47 large extrinsic loop (Putnam-Evans & Bricker, 1994, 1997; Clarke & Eaton-Rye, 1999). These results were significant because the R320S mutation represented the first site in the CP43 protein that alters the chloride requirement of PSII and inferred that the large extrinsic loop of CP43

is involved in chloride binding/sequestration at the PSII active site.

The phenotype of R320S closely resembles that of mutants in which the extrinsic cytochrome c550 protein has been deleted. Cytochrome c550 appears to regulate the efficiency of the S1 S2 and/or the S2 S3 state transitions. The *psb*V deletion strains exhibit a stabilization of the S2 state and cannot grow photoautotrophically under chloride- or calcium-limiting conditions (Shen et al., 1998). Cytochrome c550 functions in a manner similar to that of the PsbP protein in higher plant PSII. We recently have shown that R320S also does not grow under conditions of calcium depletion (Putnam-Evans, unpublished). In order to assess whether or not cytochrome c550 is associated with PSII in the mutant, a histidine-tagged version of the R320S mutant was produced to facilitate the isolation of PSII particles. These particles were analyzed for the presence of cytochrome c550 (Bricker et al., 2002). Reduced minus oxidized difference spectroscopy and chemiluminescence staining on western blots indicated that cytochrome c550 was absent in these PSII particles. Whole cell extracts from the R320S mutant, however, contained a similar amount of cytochrome c550 to that observed in the control strain. These results indicate that the mutation R320S in CP43 prevents the strong association of cytochrome c550 with the PSII core complex. Thus, the Arg320 residue may be involved in the formation of the binding domain for the cytochrome. This is the first

showed that it contained a short deletion within the *psb*C gene. Thylakoid membranes isolated from this mutant had decreased levels of the reaction center protein D2.

In studies specifically targeting the large extrinsic loop E of CP43, eight short deletions were produced in *Synechocystis* CP43 (Kuhn & Vermaas, 1993). Significantly, all resulting mutants showed complete loss of photoautotrophic growth and the ability to evolve oxygen. These mutants also contained decreased levels of the PSII reaction center proteins D1, D2 and CP47. This work clearly established a role for the large extrinsic loop in PSII function.

Site-directed mutagenesis of the large extrinsic loop of CP43 in *Synechocystis* has revealed a number of amino acids required for normal PSII function. In general, these mutants segregate into two groups. One group contains mutations at residues that appear to be involved in manganese ligation and catalysis. These sites are Glu354 and Arg357. Glu354 and Arg357 lie, respectively, within the first and second coordination spheres of the Mn4CaO5 cluster. Glu354, as stated above, provides bidentate ligation of Mn3, and also binds one chloride ion. The effects of mutation of this residue have been examined in three independently constructed mutants in which the glutamate was replaced with glutamine (Rosenburg et al., 1999; Strickler et al., 2008; Shimada et al., 2009; Service et al., 2011). The E354Q mutant grows slowly and exhibits only 20% of control rates of oxygen-evolving activity. The PSII centers of this mutant can advance to the S2 and S3 states, but do not advance past S3 (Shimada et al., 2009; Service et al, 2011). Glu354 appears to ligate manganese bound to a substrate water molecule (Shimada et al., 2009; Service et al, 2011) that undergoes rapid exchange in the S3 state (Service et al., 2011). Arg357 lies in the second coordination sphere of the Mn4CaO5 cluster. In the crystal structure, one guanidinium nitrogen of this residue is hydrogen-bonded to two of the -oxo bridges (O2 and O4) and the other is hydrogen-bonded to two residues of the D1 protein, Asp170, which ligates Mn4 and the calcium ion, and Ala344, which ligates Mn2. Mutation of this arginine to serine abolished oxygen-evolving activity and photoautotrophic growth (Knoepfle et al., 1999; Ananyev et al., 2005). A more conservative replacement of the arginine with lysine (Hwang et al., 2007), also produced a mutant that failed to grow photoautotrophically but retained a low rate of oxygen-evolving activity. S-state cycling defects were observed in this mutant. Specifically, the strain exhibited a high miss factor and a decreased yield of the S2 state. Using isotopic labeling coupled with FTIR spectroscopy, structural coupling between an arginine and the Mn4CaO5 cluster was observed (Shimada, et al., 2011). The arginine was presumed to be Arg357 of CP43, since this is the only arginine residue in close proximity to the Mn4CaO5 cluster. The data are consistent with a role for Arg357 in the structural stabilization of the cluster and/or possibly in abstracting protons during water oxidation, as has been proposed previously (McEvoy & Brudvig, 2006). These experimental data are consistent with the crystallographic data and with a direct role of both Glu354 and Arg357 of CP43 in the water splitting mechanism.

Of particular interest is a second group of CP43 mutants showing defects associated with the chloride requirement of PSII. These are mutations at Arg320 and Asp321. We have extensively characterized a *Synechocystis* mutant in which Arg320 was replaced by serine (Knoepfle et al., 1999; Young, et al., 2002). When grown in complete BG-11 media (480 M chloride), the R320S mutant strain exhibited photoautotrophic growth rates comparable to the control strain and assembled approximately 70% of the PSII centers found in the control strain (Knoepfle et al., 1999). These centers are capable of evolving oxygen to 60-70% of

showed that it contained a short deletion within the *psb*C gene. Thylakoid membranes

In studies specifically targeting the large extrinsic loop E of CP43, eight short deletions were produced in *Synechocystis* CP43 (Kuhn & Vermaas, 1993). Significantly, all resulting mutants showed complete loss of photoautotrophic growth and the ability to evolve oxygen. These mutants also contained decreased levels of the PSII reaction center proteins D1, D2 and

Site-directed mutagenesis of the large extrinsic loop of CP43 in *Synechocystis* has revealed a number of amino acids required for normal PSII function. In general, these mutants segregate into two groups. One group contains mutations at residues that appear to be involved in manganese ligation and catalysis. These sites are Glu354 and Arg357. Glu354 and Arg357 lie, respectively, within the first and second coordination spheres of the Mn4CaO5 cluster. Glu354, as stated above, provides bidentate ligation of Mn3, and also binds one chloride ion. The effects of mutation of this residue have been examined in three independently constructed mutants in which the glutamate was replaced with glutamine (Rosenburg et al., 1999; Strickler et al., 2008; Shimada et al., 2009; Service et al., 2011). The E354Q mutant grows slowly and exhibits only 20% of control rates of oxygen-evolving activity. The PSII centers of this mutant can advance to the S2 and S3 states, but do not advance past S3 (Shimada et al., 2009; Service et al, 2011). Glu354 appears to ligate manganese bound to a substrate water molecule (Shimada et al., 2009; Service et al, 2011) that undergoes rapid exchange in the S3 state (Service et al., 2011). Arg357 lies in the second coordination sphere of the Mn4CaO5 cluster. In the crystal structure, one guanidinium nitrogen of this residue is hydrogen-bonded to two of the -oxo bridges (O2 and O4) and the other is hydrogen-bonded to two residues of the D1 protein, Asp170, which ligates Mn4 and the calcium ion, and Ala344, which ligates Mn2. Mutation of this arginine to serine abolished oxygen-evolving activity and photoautotrophic growth (Knoepfle et al., 1999; Ananyev et al., 2005). A more conservative replacement of the arginine with lysine (Hwang et al., 2007), also produced a mutant that failed to grow photoautotrophically but retained a low rate of oxygen-evolving activity. S-state cycling defects were observed in this mutant. Specifically, the strain exhibited a high miss factor and a decreased yield of the S2 state. Using isotopic labeling coupled with FTIR spectroscopy, structural coupling between an arginine and the Mn4CaO5 cluster was observed (Shimada, et al., 2011). The arginine was presumed to be Arg357 of CP43, since this is the only arginine residue in close proximity to the Mn4CaO5 cluster. The data are consistent with a role for Arg357 in the structural stabilization of the cluster and/or possibly in abstracting protons during water oxidation, as has been proposed previously (McEvoy & Brudvig, 2006). These experimental data are consistent with the crystallographic data and with a direct role of both Glu354 and Arg357

Of particular interest is a second group of CP43 mutants showing defects associated with the chloride requirement of PSII. These are mutations at Arg320 and Asp321. We have extensively characterized a *Synechocystis* mutant in which Arg320 was replaced by serine (Knoepfle et al., 1999; Young, et al., 2002). When grown in complete BG-11 media (480 M chloride), the R320S mutant strain exhibited photoautotrophic growth rates comparable to the control strain and assembled approximately 70% of the PSII centers found in the control strain (Knoepfle et al., 1999). These centers are capable of evolving oxygen to 60-70% of

isolated from this mutant had decreased levels of the reaction center protein D2.

CP47. This work clearly established a role for the large extrinsic loop in PSII function.

of CP43 in the water splitting mechanism.

control rates at saturating light intensities. However, they show an enhanced susceptibility to photoinactivation. Additionally, the fluorescence rise time for this mutant was increased by a factor of two over the control. Thus, the PSII centers in this mutant do not function normally, even under normal growth conditions. Significantly, when the R320S mutant was grown in media depleted of chloride (30 M chloride), it exhibited a severely reduced photoautotrophic growth rate. The effect of chloride depletion on the growth rate of the mutant was reversed by the addition of 480 M bromide to the chloride-depleted BG-11 media. It is well known that bromide will functionally replace chloride at the oxygenevolving site in cyanobacteria grown on bromide-containing media (Yachandra et al., 1993). Additionally, it is well known that bromide can replace chloride in supporting oxygen evolution in isolated thylakoids (Kelly & Izawa, 1978) and PSII membranes (Sandusky & Yocum, 1983). Oxygen evolution rates for the mutant were further depressed to about 22% of the rate observed in control cells under chloride-limiting conditions. Addition of bromide restored these rates to those observed under chloride-sufficient conditions. The mutant exhibited a significantly lower relative quantum yield for oxygen evolution than did the control strain and this was exacerbated under chloride-limiting conditions. Fluorescence yield measurements indicated that both the mutant and control strains assembled fewer PSII reaction centers under chloride-limiting conditions. The reaction centers assembled by the mutant exhibited an enhanced sensitivity to photoinactivation under chloride-limiting conditions, with a t1/2 of photoinactivation of 2.6 min under chloride-limiting conditions compared to a t1/2 of 4.7 min under normal growth conditions. The mutant also exhibited an enhanced stability of its S2 state and an increased number of centers in the S1 state following dark incubation. These results indicate that the mutant R320S exhibited a defect in its ability to utilize chloride in support of efficient oxygen evolution in PSII. This was the first mutant of this type described in the CP43 protein**.** In terms of this chloride effect, R320S has a similar phenotype to the R448G, K321G, and F363R mutants constructed in the CP47 large extrinsic loop (Putnam-Evans & Bricker, 1994, 1997; Clarke & Eaton-Rye, 1999). These results were significant because the R320S mutation represented the first site in the CP43 protein that alters the chloride requirement of PSII and inferred that the large extrinsic loop of CP43 is involved in chloride binding/sequestration at the PSII active site.

The phenotype of R320S closely resembles that of mutants in which the extrinsic cytochrome c550 protein has been deleted. Cytochrome c550 appears to regulate the efficiency of the S1 S2 and/or the S2 S3 state transitions. The *psb*V deletion strains exhibit a stabilization of the S2 state and cannot grow photoautotrophically under chloride- or calcium-limiting conditions (Shen et al., 1998). Cytochrome c550 functions in a manner similar to that of the PsbP protein in higher plant PSII. We recently have shown that R320S also does not grow under conditions of calcium depletion (Putnam-Evans, unpublished). In order to assess whether or not cytochrome c550 is associated with PSII in the mutant, a histidine-tagged version of the R320S mutant was produced to facilitate the isolation of PSII particles. These particles were analyzed for the presence of cytochrome c550 (Bricker et al., 2002). Reduced minus oxidized difference spectroscopy and chemiluminescence staining on western blots indicated that cytochrome c550 was absent in these PSII particles. Whole cell extracts from the R320S mutant, however, contained a similar amount of cytochrome c550 to that observed in the control strain. These results indicate that the mutation R320S in CP43 prevents the strong association of cytochrome c550 with the PSII core complex. Thus, the Arg320 residue may be involved in the formation of the binding domain for the cytochrome. This is the first

Mutations in the CP43 Protein of Photosystem II

temperature overnight. Afterward, the DNA was stored at 4C.

elongation step each cycle, and Step 4 – 72C for 7 minutes.

at 72C for 10 minutes.

Affect PSII Function and Cytochrome C550 Binding 61

was added and the mixture was incubated at 37C for 15 minutes. Then, 20 l of 500 mM EDTA, 50 l of proteinase K (10 mg/ml in 50% glycerol), and 55 l of 10% sarkosyl were added. The mixture was then incubated at 55C for 15 minutes. After 5 minutes at room temperature, 600 l of TE (5 mM Tris-HCl, pH 8.0, 5 mM EDTA) saturated phenol was added and the mixture incubated at room temperature for 10 additional minutes with gentle inversion. The mixture was then centrifuged at 18000 x *g* for 3 minutes and the aqueous phase was transferred to a new tube. Next, 100 l of 5 M NaCl, 100 l of 10% CTAB, and 600 l of chloroform were added to the sample. The tubes were then shaken on a rotary shaker for 15 minutes, centrifuged at 8300 x *g* for 3 minutes, and the aqueous phase was transferred to a new tube. Next, an equal volume of cold 100% isopropanol was added to the sample, and the tubes were gently inverted. The mixture was allowed to incubate at room temperature for 20 minutes to precipitate the DNA before centrifuging the sample at 8200 x *g* for 10 minutes at 4C to pellet the DNA. After centrifugation, the supernatant was removed and discarded, and 1 ml of cold 70% ethanol was added to the tube to wash the DNA pellet. The tubes were then centrifuged again for 10 minutes at 4C. Following this, the supernatant was quickly removed and the DNA pellet was allowed to air dry. Then, 100 l of TE buffer was added to the tube and the DNA was allowed to resuspend at room

The *psbC* gene was amplified by PCR. A typical reaction consisted of 79.7 l of sterile dH2O, 10 l of 10X PCR buffer (Invitrogen, Inc.), 4 l of 50 mM MgCl2, 0.8 l of 10 mM dNTPs, 2 l of forward primer (2 pmol/l) (Invitrogen, Inc.), 2 l of reverse primer (2 pmol/l) (Invitrogen, Inc.), 30-60 ng of genomic DNA, and 0.5 l of Taq DNA polymerase (2.5 units, recombinant). The thermal cycling routine consisted of the following steps: Step 1 – 1 cycle of 94C for 3 minutes, 60C for 45 seconds, and 72C for 2 minutes, Step 2 – 25 cycles of 94C for 1 minute, 60C for 45 seconds, and 72C for 2 minutes, Step 3 – 5 cycles of 94C for 1 minute, 60C for 45 seconds, and 72C for 2 minutes with a 5 second extension added to the

Samples were cleaned using a QIAquick PCR Purification Kit (Qiagen, Inc.) and resuspended in 40 l of sterile dH2O. Following the cleanup, the samples were subjected to sequencing reactions in a PTC-100TM Programmable Thermal Cycler (MJ Research, Inc.). The components of these sequencing reactions were 5 l of sterile dH2O, 2 l of Big Dye (Applied Biosystems, Inc.), 3 l of Big Dye reaction buffer, 2 l of primer (2 pmol/l), and 30-60 ng of purified PCR product DNA. The cycling routine was as follows: 26 cycles of 94C for 1 minute, 55C for 1 minute, and 72C for 2 minutes, followed by a final elongation

After the completion of the sequencing reactions, the DNA was alcohol precipitated with ethanol and was resuspended in a 5:1 formamide:50 mM EDTA/50 mg/ml Blue Dextran mixture. Then, samples were loaded in an ABI Prism 377 DNA Sequencer (Applied Biosystems, Inc.) and the sequencer was run for 7 hours. Sequences were analyzed using the

Oxygen evolution activity was measured by O2 polarography with a Hansatech oxygen electrode (Knoepfle et al., 1998). Assays were performed at 25°C on whole cells in complete or chloride-depleted BG-11 media with 1 mM 2,6-dichlorobenzoquinone (DCBQ) added as

Auto Assembler DNA sequence analysis software (Applied Biosystems v.1.4.0).

**5.3 Oxygen evolution assays and photoinactivation assays** 

residue identified in any PSII protein that potentially provides a binding site for cytochrome c550. We have also begun preliminary investigation of mutants produced at the residue adjacent to Arg320, Asp321. Replacement of this aspartate by asparagine produced a mutant with a phenotype very similar to that of the R320S mutant (Putnam-Evans, unpublished). That is, this mutant exhibits almost normal PSII function when grown in complete BG-11 media, but fails to grow photoautotrophically under chloride-limiting conditions and exhibits a marked decrease in oxygen evolution rates under chloride depletion.

Examination of the newest crystal structure reveals that Arg320 lies at the interface between cytochrome c550 and PsbU, within 2.9 angstroms of Asn49 of cytochrome c550, close enough to form a hydrogen bond. Another potential hydrogen bonding partner in *Synechocystis* is Asn51. However, in *Thermosynechococcus*, the organism from which the x-ray structure is derived, this residue is a serine residue. Nevertheless, it is located only 3.6 angstroms from Arg320. Asp321 lies within 2.7 angstroms of Asn99. Additionally, Arg320 and Asp321 of CP43 lie within 2.8 angstroms of each other. Thus, these sites are candidates for potential ligands to cytochrome c550 and PsbU. It is tempting to speculate that Arg320 of CP43 is a central player in these interactions, perhaps via a hydrogen bonding network involving several of these and perhaps other residues. Here we report the construction and preliminary characterization of two additional mutations at site 320, R320D and R320K, towards the goal of better understanding the role of this amino acid in these potential protein-protein interactions in PSII.
