Secondary Protein Structure and Pattern Database Predictions of MSP-

The MSP-7 amino acid sequence was entered into a variety o f secondary

structure prediction programs and used to search pattern databases to identify conserved

motifs present in MSP-7.

To determine whether the MSP-7 precursor was targeted to the merozoite surface

by a signal sequence and if MSP-7 had any transmembrane domains, the 351 residue

sequence was submitted to signal sequence and trans-membrane prediction programs.

Psort II and SignalP, using networks trained on eukaryotic data, gave differing

results. Psort II, using a combination o f PSG and von Heijne’s method, did not predict a

N-terminal signal peptide, instead using the ^-nearest neighbour (A:-NN) algorithm to

predict the following probabilities o f location for MSP-7; 34.8 % nuclear, 21.7 %

extracellular / cell surface, 17.4 % mitochondrial, 13.0 % cytoplasmic, 8.7 % vacuolar,

4.3 % vesicles o f secretory system. However, Signal P predicted a putative signal

sequence o f 27 amino acids, with the most likely cleavage site between residues 27 (Ser)

and 28 (Thr) (Figure 3.8). Until a signal prediction program trained on protozoan or

Plasmodium signal sequences is available, and since MSP-722 has been located on the

merozoite surface suggesting it is targeted to the plasma membrane (McBride and Heidrich 1987; Stafford et ah 1996), the Signal P prediction can be assumed to be

correct. Similarity o f the predicted MSP-7 signal sequence with other malarial surface

proteins associated with the MSP-1 complex, such as MSP-6 remains to be elucidated.

TMpred and DAS transmembrane prediction servers denominate a possible

transmembrane helix between residues 2 to 21. Assuming MSP-7 has a signal sequence

o f 27 residues and not a transmembrane domain at the N-terminus, the resulting 324

amino acid protein would have a predicted molecular mass o f 38.5 kDa and an estimated

pi o f 4.52.

3,6.1 Secondary Protein Structure

A combination o f the Chou-Fasman and Robson-Gamier secondary structure

prediction programs available on Mac Vector 6.5, and the Profile fed neural network

Chapter Three : The Merozoite Surface Protein - 7 Gene and Protein 133

predict the secondary structure and solvent accessibility for iÿMSP-7. As no homologues

were identified in SWIS-PROT the expected accuracy o f the PHD prediction programs

was only between 62 - 6 6 %. The secondary structure o f MSP-7 is predominantly alpha

helical with two small regions o f beta sheet and several beta turns indicating the protein

is highly fiexible (See Figure 3.12).

SMART sub-program SEG predicted that residues: 7 - 20, 37 - 52, 62 - 75, 92 -

113, 328 - 339 are o f low compositional complexity whilst Prospero predicted the

presence o f two internal repeats between residues 117 - 252, and 125 - 262.

MSP-7 is mainly hydrophilic, with a negative charge cluster at residues 94 - 148,

which may form the core backbone o f the structure. Psort II predicted a coiled-coil

region fi*om residues 234 (Lys) to 261 (Tyr) with a hydrophobic region in the coil

perhaps a putative region for interaction with MSP-1.

3.6.2 Pattern Database Searches

The MSP-7 amino acid sequence was used as the query sequence to search a

number o f secondary or pattern databases for conserved motife that could give an

indication o f function for MSP-7.

PHD and ScanProsite searches o f the PROSITE database predicted the presence

in MSP-7 o f a number o f patterns including N-glycosylation, kinase phosphorylation and

N-myristoylation sites, although these are probably felse positives. Additionally a

PrintsScan search o f the PRINTS fingerprint database predicted the presence o f

tropomyosin (residues 6 6 - 84, 229 - 258, and 332 - 358), and alpha-tubulin (residues

236 - 249, and 2 9 9 -3 1 5 ) motifs in MSP-7.

A PRODOM domain search performed by PHD recovered no significant

homologous domains, whilst a Pfem search for Hidden Markov Models and a NCBI-CD

BLOCKS + and PRINTS 23.0 databases for similarity to highly conserved aligned

motifs but no significant matches were obtained.

These results do not predict a known function for MSP-7. This may be because

the motifs present in MSP-7 remain to be discovered or that the programs used in the

analysis are not sophisticated or comprehensive enough to deal with protozoan motifs.

3.6.3 Sequence Conservation Between / y MSP-7 Processing Sites and Other Merozoite Surface Proteins

The P f MSP-722 cleavage site shows similarity to protease cleavage sites at the

start o f MSP-130 (MSP-1 primary processing site) and MSP-636 (Figure 3.13A). This

suggests that the protease responsible for these cleavages has a similar specificity. Since

the cleavages are all predicted to occur at the same stage o f merozoite development and

the C-terminal products are known to be complexed together in the shed MSP-1

complex, it is possible that the same protease is responsible for these cleavages. Once an inhibitor o f the protease SERPH that is predicted to be responsible for primary MSP-1

processing is discovered, this hypothesis can be tested (Blackman 2000).

The P f MSP-722-19 N-terminal sequences, especially the N-terminal region of

MSP-7i9, show similarity to a motif in the MSP-2 gene family (Figure 3.13B). This

requires further analysis o f MSP-2 processing and binding sites before any hypothesis

concerning this motif can be made with confidence.