Although our publication of the discovery of the two peptides we named the hypocretins did not occur until 1998, the road to their discovery began in the spring of 1979. One of us (JGS) was completing postdoctoral studies with Richard Lerner at the Scripps Clinic and Research Foundation (now The Scripps Research Institute) following doctoral work performed under Wally Gilbert at Harvard, where the main thesis efforts had been in scaling up the DNA sequencing procedures developed by Gilbert and Allan Maxam. Those studies, which were described in a Reflections piece in TIBS,1 represented the first time DNA analysis was used to determine the sequence of a protein in the absence of peptide sequence information from the protein itself. Thus, one learned first hand about the superior accuracy and rapidity of carefully collected DNA-based sequence information and the possibility that one could use the methods to determine the sequences of previously unrecognized genes for which neither proteins nor RNAs were known. This was a quarter of a century before mammalian genome sequences began to appear.
In Lerner's laboratory, JGS was working with Tom Shinnick on determining the first retrovirus genome sequence, that for Moloney murine leukemia virus.2 Lerner knew of his long-term interest in neurobiology. The eminent neurobiologist Floyd Bloom, then a professor at the neighboring Salk Institute, asked Lerner if the technologies JGS had brought to Scripps could be used to study the brain. Bloom had just read the report describing the cDNA cloning of proopiomelanocortin (POMC),3 and was particularly interested in finding out about undiscovered peptide neurotransmitters. Lerner passed alog the question.
JGS answered that, since so little was known about the molecular operation of the brain, a rational approach would be to construct cDNA libraries from brain mRNA. Individual cDNA clones could then be isolated from such libraries and their nucleotide sequences determined, thus allowing the amino acid sequence of the protein encoded by the corresponding brain mRNA to be conceptually translated. cDNA cloning had recently been developed, and it was the route by which clones were obtained for mRNAs encoding particular, already known proteins, such as POMC. cDNA libraries represent all of the mRNAs expressed in the tissue from which the sample was isolated and, thus, such libraries could inform us about the complete protein set, including those proteins that were not yet identified. But that led to an obstacle: how could we use the conceptual protein sequences translated from the cDNA sequences to learn about the putative proteins themselves? The proteins would not have been seen previously, and there were few protein sequences to which to compare the new sequences. The solution was to prepare synthetic peptide fragments of the proteins and use these to elicit antibodies that would react with both the peptides and the novel protein itself, thus facilitating its biochemical and anatomical characterization. The first opportunity to apply this approach on a putative protein in the Moloney virus sequence.4
Bloom, his then-postdoctoral fellow Rob Milner and JGS began a collaboration to test these ideas. The first effort was to use northern blot hybridization to characterize the mRNAs corresponding to clones in a brain cDNA library. By analyzing the size, abundance and tissue distributions of the mRNAs corresponding to nearly 200 clones isolated randomly from a rat brain cDNA library,5 the team calculated that the 108 to 2x108 nucleotides of mRNA complexity expressed by brain corresponded to 20,000 to 40,000 distinct mRNAs, numbers that compare favorably with modern estimates since entire genome sequences have been solved. Of these, approximately 65% were enriched in the brain compared to peripheral tissues. Most were of low abundance, on the order of one part in 105. The team raised antisera directed against synthetic peptides corresponding to one of the first partial putative brain protein sequences determined, and used these to detect the protein in brain extracts and to conduct a preliminary anatomical description of the protein later shown to be myelin-associated glycoprotein.6
These early studies represent the beginning of what have since come to be known as open-system approaches to mRNA expression analysis: mRNAs are detected because of their property of being expressed in the tissue sample isolated for study. Refinement of this approach led to the discovery of the hypocretins. The notion that the tools of molecular biology could be used to address fundamental questions about the operation of the mammalian brain was controversial at the time, but is one that no one today would argue. From the sequences of brain cDNAs, we learn about the nature of brain proteins. The cDNA clones also allow the study of the brain genes themselves and, with the advent of transgenic and knockout technologies, allow the power of genetic analysis to be brought to bear on the central nervous system, permitting a forceful molecular dissection of CNS physiology.
Was this article helpful?