Proteolysis of [’ 4C]A~idoatra~ine-Labe11ed Peptides of Pea Thylakoid Photosystem 2

Julian Whitelegge,” Philip Jewess,h Patrick Camillerih & John R . Bowyer”

“Biochemistry Department, Royal Holloway and Bedford New College. Eghani Hill. Egham. Surrey TWZO OEX. U K hShell Research Ltd. Sittinghourne, Kent, ME9 8AG, U K

Many classes of herbicides competitively displace plastoquinone from t he Q B

binding site on the D1 polypcptide of Photosystem 2 (PS2). thereby inhibiting photosynthesis. Early studies using photoaffinity labels and sequence analysis of herbicide-resistant mutants, combined with the homology between PS2 and the resolved structure of the purplc photosynthetic bacterial reaction centre, suggested that the QB pocket lies between the putative 4th and 5th membrane spanning alpha helices of D1. However, many classes of commercial PS2 herbicides, such as the phenylureas and phenols, do not bind to the bacterial reaction centre, indicating a somewhat different structure for the PS2 herbicide binding site. Photoaffinity analogues of the triazine and phenylurea classes of herbicides have therefore been used to study the molecular architecture of the QB site of PS2.

The strategy used involves the photoaffinity labelling of isolated thylakoids followed by proteolysis and isolation of labelled peptidcs, the aim being to sequence such pcptides to reveal amino acids derivatised by the photoaffinity label. lise of photoaffinity analogues of different herbicides, or use of similar herbicides with the photoactive group in different positions. will help map the Q B site. [ ‘‘C]azidoatrazine (2-azido-4-ct hylamino-6-isopropylamino- 1,3,5-tria~ine; azidoatrazine is not an ISO/BSI approved common name) is used to develop protocols for this work.

Photoaffinity labelling of isolated thylakoids with [ “C]azidoatra7ine results in labelling of the 32 kD D1 polypeptide’ and it is necessary to separate this polypeptide from other thylakoid polypeptides in this region. This has been achieved by controlled solubilisation of labelled Pisirm S U ~ ~ L ~ L M I (cv Feltham First) thylakoids with ‘Triton X-100’ and isolation of PS? reaction centres using ion exchange chromatography.2 Fluorography after sodium dodecylsulphate- polyacrylamide gel electrophoresis (SDS-PAGE) of labelled reaction centres revealed labelled bands at 32 kD and 55 kD of approximately equal abundance. As the 55 kD band was not seen in isolated thylakoids it seemed likely that this band represented a dimerisation product of the 32 kD polypeptide. Studies with antibodies and peptide mapping3 indicated that the 5 5 kD band was, in fact, a heterodimer of the D1 and the D2 polypeptides, emphasising their close proximity to each other in the PS2 reaction centre. It is not known whether D1D2 heterodimer formation is a random process or whether a heterogeneity in the interaction between D1 and D2 within PS2 reaction centres results in preferential monomer or heterodimer formation during SDS-PAGE. The ‘‘C-labelling site in isolated

labelled DI and DI D2 heterodimer is being analysed to investigate the possibility that there is heterogeneity of ['4C]azidoatrazine binding.

Labelled DI monomer and DI D2 heterodimer were isolated by electroelution from polyacrylamidc gels after SDS-PAGE. The bands, identified by niolccular weight after staining, were excised and soaked in 0. I M Tris + 0.1 M Tricine + 1 g litre-' SDS (pH 8.25) for 5 min and then placed within a piece of dialysis tubing (SpectraiPor 6. molecular weight cut off, 1000 kD). After electroelution overnight i n a 70 V electric field (tubing perpendicular to field) i n the same buffer, the gel was removed from the tubing. The contents of the tubing were then dialysed for 5 days against distilled water to remove as much buffer and SDS as possible prior to freeze drying. Peptide isolated in this manner was then treated with either trypsin or Strrph~~/ococ.c~~r.s V 8 protease in ammonium bicarbonate (100 niv: pH 7.8). The substrate/enzyme mixture (approximate molar ratio, 50:l) was incubated at 37 'C for 3 h and the resulting mixture analysed by further SDS-PAGE.

The results showed that the DID2 heterodimer was stable during re- electrophoresis but after treatment with trypsin or V 8 protease it was degraded, resulting in a 20 kD labelled fragment. Proteolysis of labelled DI, isolated by electroelution. resulted in a similar 20 kD labelled fragment. This peptidc was resistant to further degradation by either protease.

These data localise the [L4C]azidoatrazine binding site to a similar region of D I and attempts are being made to identify labelled amino acid residues via protein sequencing. The labelling of the 20 kD fragment is consistent with the reported labelling site in the region Met 214 to Arg 22j4 and proteolysis of labelled D1 at an exposed region of the polypeptide between Arg 225 to Glu 244.

PS2 heterogeneity is being further invcstigated using EPR techniques and biochemical studies are being conducted to investigate whether post-translational modifications. such as phosphorylation, might give rise to such heterogeneity.

R< ferences I . Pfister, K. , Steinback, K . E., Gardner, G. & Arntzen, C. J . , Proc. ~ V t r t . A d . S c i . U S A , 78

2. Barber, J . , Chapman, D. J . & Telfer. A., F E B S Lctt., 220 ( 1987) 67-73. 3. Marder, J . B., Chapman. D. J.. Telfer. A.. Nixon, P. J. & Barber. J., Pltrnt Mol. Hiol . . 9

(1987) 325-53. 4. Wolber, P. K.. Eilmann, M. & Steinback, K . E., Arch. Biocheni. Biophj~s., 248 (1986) 224-

33.

(1981) 981-5.

Patch Clamp Studies of Insect 4-Aminobutyric Acid Receptors

Alison R. Taylor 6i David J. Beadle

Department of Biological and Molecular Sciences. Oxford Polytechnic. Oxford, OX3 OHP. CIK

Neuropharmacological studies on insect neuroncs have shown that they possess a receptor to the inhibitory neurotransmitter 4-aminobutyric acid (GAHA).' The

experiments described here involve the use of the patch clamp to make elcctrophysiological recordings from adult and cultured insect neurones and to obtain more information about the insect GA RA receptor-ionophore complex. The responses to GABA and a variety of putative agonists were recorded in ‘whole cell’ and ‘patch’ configurations. Agonist-induced current fluctuations recorded in whole-cell mode were processed using fluctuation analysis.

Adult neurones from the brains of Locirstu niiyrutorin were dissociated and maintained in ‘short term culture’. Supraoesophageal ganglia were dissected from the adult locusts and digested in 002 ?{, collagenase (Sigma type l a ) for 40 min at 37°C. The ganglia were then aspirated through the tip of a Pasteur pipette to dissociate the cells. The resulting neuronal suspension was spun down (< 8009) and washed in fresh saline (210 mM sodium chloride, 10 mM calcium chloride, 3.1 mbi potassium chloride, 10 m M Tris buffer at pH 7.2) at least three times. The final pellet of cells was resuspended in 5 + 4 culture media and thcn plated out on to Petri dishes.* Locust neurones maintained in this way were viable for recording for 48 h: they varied in size between 10-40pm, the smaller cells being selected for electrophysiology. After several hours, growth cones could be observed. The resting potential of the cells varied from - 28 mV to - 45 mV, the average being - 35 mV. The time constant was 10-20ms and input resistance 50-150 MR. All cells exhibited rectification on passing a depolarising current pulse and in a few, action potentials could be elicited. The preparation and properties of the cockroach cultured neurones have been described previously.’

Cells were allowed to equilibrate in saline before recording commenced. The patch electrode contained 114 mM potassium chloride: 5 mM ethylenedioxybis (ethylencnitrilo) tetra (acetic acid) (EGTA); 1.6 mM magnesium chloride; 0.2 mM calcium chloride: 80 mM 1)-glucose and 10 mM Tris buffer pH 7.2. GABA and putative agonists were made up in saline M ) and applied to the cells using pressure ejection from a micropipette. The chloride channel blocker and GABA receptor antagonist picrotoxin (PTX) was made up in saline (lo-’ M ) and applied by bath perfusion. Voltage clamp recordings were made using a List EP7 or Dagan patch clamp amplifier; the clamp potential in all experiments was - 60 mV. Current fluctuations recorded during prolonged application of agonist were processed using noise ana ly~ i s .~ The half-power frequency and variance of the noise were thcn used to calculate single-channel open time and conductance values for the agonists. Single-channel recordings were obtained by applying agonist to out-side out patches.

The results for the fluctuation analysis of GABA and the putative agonists are shown in Table 1. In all cases the spectra obtained could be reasonably fitted by a single Lorentzian component, suggesting that a single class of ion channel was involved in the agonist evoked response. All compounds tested on the locust and cockroach neurones induced a hyperpolarising response, accompanied by an increase in conductance of the cell. PTX blocked the responses in all cases except isonipecotic acid, where it was not tested. The reversal of the responses was between - 20 mV and 0 mV, the chloride cquilibrium potential being calculated as - 11 mV. The mean channel open time for GABA in the locust adult cells was 12.9 (k 2.9) ms and was significantly different (95% Student’s t test) from the open time in the

TABLE I Noise Analysis of Agonist-Induccd Current Fluctuations in Locurtu migrutoriu Adult Neurones Maintained in Short Term Culture and Periplurwta uriiericunu Embryonic Cultures

Agolli.sr Nu,,1her qf cells

__ Locusttr rnigrtctorirc adult cells GAHA 5 ZAPA 7 Muscimol 2 Isoguvacine 7 Isoriipecotic acid 5

Periplu~iotu uniericmiu embryonic cultures GAHA 10 Musciinol 4

12.9 (k 2.9) 11.1 ( k 2 . 4 ) 17.9 ( f 2 . 8 ) 6.0 [ k 1.3) 8.6 ( 2 3.6)

26.1 (k6.5) 14.5 (i 1.7)

21.9 ( k 5 . 3 )

12.x (k0.7) 17.3 ( & 5 6 )

19.1 ( k 5 . 9 ) 14.9 [ 7.5)

18.8 (2 5.9) 18.5 ( 6.2)

2 *i

i

T J T

10-2 I I I I 4 + t + t + t t !

;o-l 2 4 6 BlAO 2 4 6 B;o1 2 4 6 8102 2 4 6 8103

Frequency ( H z )

Fig. I . (a ) Power spectrum of the noise induced by application of G A B A onto a cockroach ncurone at -60 mV. The spectrum is fitted with a single Lorentzian curve with a corner frequency of 7.5 Hz. This gives a value of 21 tns for mean channel open time. (b ) Representative trace of GABA-evoked channels in locust neurones. G A B A wasapplied to theoutsideofan out-side out patch at -60 mV. Calibration bars

are 10 ms (horizontal) and 2 pA (vertical).

cockroach cultured cells, where the value was 26.1 ( t 6.5) ms; the single channel conductance of 21.9 ( k 5 3 ) p S and 18.9 ( k 5 9 ) p S for locust and cockroach respectively were not significantly differcnt. Isoguvacine had the same conductance value as the other compounds tested on the locust ncurones though it produced the significantly shorter open time 6.0 (k 1.3) ms. Single channel recordings from out-

side out patches have been obtained for GABA, 3-[(aminoiminomethyl)thio] propenoic acid (ZAPA) and muscmol and these are currently being analysed. A typical record of single channel openings after application of GABA is shown in Fig. 1 . The single channel values reported here suggest that there are subtle differenccs in the pharmacology of the insect GARA receptor. which may be due to the different species or different tissue used.

Rejirences I . Lees, G.. Beadle, D. J., Neumann, R . & Benson, J. A,, Brctrri Res., 401 (1987) 776 8 2 . Lees, G., Beadle. D. J., Botham, B. P. & Kelly, J . S.. J . Itisecr I'lzysiol., 31 (1985) 135 44. 3. Neher, E. & Stcvcns, C. F., A m . Ror. Hiophy.5 Bioeriy., 6 (1977) 345-81.