Telomerase - Overview

CAMBIA owns patents on splice variants of transcripts of the gene encoding human telomerase, important in regeneration of chromosomes during successive cell divisions.  There is much interest in exploiting this gene for cancer diagnostics, stem cell uses, anti-aging research etc.

About this technology landscape

• Originally authored by Carol Nottenburg (PhD, JD).  Technology landscapes, by their very nature, become outdated.  We welcome updates and inputs by others through the comments interface available on every page of this version of the technology landscape.
• Production by "The Team"

Cellular senescence and telomerase

Cells of nearly every complex organism do not have an unlimited ability to divide.  This phenomenon was described by Leonard Hayflick in 1961 (Hayflick 1965).  The number of potential divisions - about fifty - was dubbed the “Hayflick Limit” and sometimes is called cellular senescence.  For the first time, it was appreciated that cells could be mortal (normal cells) or immortal (tumor cells).  This distinction underpins much of modern cancer research. 

The mechanism accounting for the Hayflick Limit remained an enigma for many years to come.  The first major clue occurred in 1986 when Cooke and Smith observed that the ends (telomeric regions) of human sex chromosomes were substantially shorter in non-germline cells than in germline cells (Cooke and Smith 1986).  Like scientists of their time, they appreciated that in the absence of an active mechanism, some telomeric DNA would otherwise be lost during each replication cycle.  Less than a year earlier, Greider and Blackburn provided evidence for an enzyme-mediated mechanism to maintain telomeres:  they discovered that the ends of the linear chromosomes in Tetrahymena, a ciliated protozoan, were maintained by an enzyme, telomerase, which added new terminal DNA sequences (Greider and Blackburn 1985).  In their paper, Cooke and Smith hypothesized that telomerase might not be active in somatic cells (non-germline cells).  Furthermore, they provocatively proposed that sustained loss of telomeric DNA could eventually limit the ability of somatic cells to divide.  Eventually a mechanism to explain the phenomenon emerged, two and a half decades after discovery of the Hayflick Limit. 

The hypothesis that telomere loss eventually limits replication of human cells was not universally accepted.  Critics cited apparent exceptions:  telomeres in mouse cells are much longer than in human cells, but mouse cells don’t have significantly more proliferative potential (Kipling and Cooke 1990); and senescent human cells still have telomeres.  On the other hand, many observations supported the causal relationship between telomerase, telomeres and cell proliferation.  For example, human cells with shorter telomeres could not replicate as many times as cells with longer telomeres (Allsopp et al. 1992), and telomerase activity was detected in immortal or tumor cells but not in normal cells (Kim et al. 1994).  While the evidence favored a causal relationship, the evidence was circumstantial or correlative. 

Definitive proof that shortened telomeres are responsible for cellular senescence finally emerged in 1998.   Bodnar et al. forced expression of telomerase in normal human cells by transfection of retinal pigment epithelial cells and foreskin fibroblasts with a vector encoding the human telomerase enzyme.  Remarkably, these cells exhibited elongated telomeres, “divided vigorously”, and proliferated at least 20 doublings beyond their normal life-span; in contrast, the control cells showed shortening of telomeres and senescence (Bodner et al. 1998).  Thus, in typical somatic cells, human telomeres normally undergo shortening at each cell division, and when several kilobases of telomeric DNA is gone, cell division halts and senescence manifests. 

Unlimited proliferative potential of cells is not necessarily advantageous to organisms.  It may seem that an unlimited capacity to replicate would be a good thing - the body could undergo self-repair in response to disease or trauma.  Yet, replication is not a risk-free event – mutations can accumulate, chromosomes can break or incompletely separate, etc.  If multiple mutations are required for tumorigenesis, then fewer replications favor maintenance of a normal phenotype.   Thus, one consequence of the Hayflick Limit appears to be beneficial to the organism, acting as part of a tumor suppressor mechanism. 

Telomerase gene structure and expression

The elusive telomerase gene was long sought after.  Candidates from a number of organisms came and went until investigators finally turned their attention to Euplotes, a hypotrichous ciliate.  These ciliated protozoans contain two types of nuclei:  micronuclei and macronuclei.  The macronucleus of hypotrichous ciliates contains at least 10 million small DNA molecules (about 1800 - 2500 bp long).  Moreover, each of the DNA molecules have telomeric sequences capping their ends.  (Hoffman et al. 1995)

Fig. 1.  Euplotes on a droplet of water.

^{Used by permission of www.microscope-microscope.org}

The ribonucleoprotein containing telomerase was purified from Euplotes macronuclei by affinity chromatography with antisense 2’-O-methyl oligonucleotides (Lingner and Cech 1996).  The active complex contained two proteins, a 123 kD and 43 kD protein, along with an RNA subunit.  The larger protein proved to be the catalytic subunit - analysis of the predicted translation of the gene sequence revealed characteristic reverse transcriptase (RTase) motifs (Lingner et al. 1997).  The requirement of the RTase motifs for telomerase activity was shown using Est2p, a yeast homolog.  Conserved amino acids in the RTase motifs of Est2p were changed by site-directed mutagenesis, and the mutants transformed into yeast lacking the est2 gene.  Expression of the mutant proteins led to senescence and shortened telomeric tracts (Lingner et al. 1997).  This provided the best evidence that the 123 kD protein from Euplotes and Est2p were the catalytic subunit of telomerase. 

Despite knowledge of the DNA and protein sequences of two telomerases, direct cloning of the gene for human telomerase was not possible.  Generally, using sequence from one species to find the homolog in an evolutionarily distant second species is most successful when used to probe cDNA libraries.  Because humans have only 46 chromosomes (92 telomeres), very little telomerase protein - and very low amounts of RNA encoding telomerase - is expected, even in immortalized cells.  Fortunately, in the late 1990s, large scale EST (expressed sequence tags) projects were underway, both in industry and in academia.  A BLAST search of EST databases with the Euplotes telomerase sequence as the query sequence identified a single EST as a significant match (Kilian et al. 1997; Meyerson et al. 1997d; Nakamura et al. 1997).  The EST sequence was then used to identify cDNA clones in libraries constructed from transformed human cell lines.  Three research groups used this approach to nearly simultaneously isolate a human telomerase cDNA (Kilian et al. 1997; Meyerson et al. 1997c; Nakamura et al. 1997).  The importance of this discovery was underscored by numerous write-ups in mainstream news organizations, such as the New York Times, the Washington Post, and Associated Press. 

Subsequently, telomerase gene sequences have been cloned from a plethora of other organisms.  The organisms include other mammals (e.g., macaque, mouse, rat, cattle, pig, dog), birds, amphibians (e.g., Xenopus laevis), echinoderms (e.g., sea urchin), insects (e.g., bee, silkworm, drosophila), Caenorhabditis elegans, Trypanosoma brucei, plants (e.g., corn, barley), and fungi (e.g., Aspergillus, Cryptococcus). 

Telomerase has reverse transcriptase motifs

The human telomerase (hTERT) sequence identified by all three groups has an open reading frame of 1132 amino acids predicted to encode a 127 kDa protein.  The identification of hTERT as a member of the telomerase family was based on the presence of motifs characteristic of the telomerase and reverse transcriptase (RT) family of enzymes.  In particular, hTERT contains six conserved motif sequences, shown below aligned to homologous motifs in other telomerase sequences and to RT motifs in HIV-1 RT (because the boundaries of these motifs are based on similarity and identity with other telomerase sequences, the functional boundary of each motif may be different.).  Consensus telomerase motif sequence is shown above and consensus RT motif sequence is shown below the alignments.  Notably, the invariant aspartic acid residues implicated in RT catalysis are conserved in the hTERT sequence. 

Fig. 2.  Alignment of RT sequence motifs of telomerase proteins.

For historical reasons, the motifs are called “1”, “2”, “A’”, “B”, “C”, and “D”. 

Although the overall homology among the telomerase proteins is relatively low (approximately 40% similarity in all pair-wise combinations), the overall structure of the protein seems to be well conserved.  Four major domains: N-terminal, basic, reverse transcriptase (RT) and C-terminal are present in all telomerase proteins.  The regions of most sequence similarity are within the RT domain and in a motif located N-terminal to motif 1 (Nakamura et al. 1997; Meyerson et al. 1997b).  This first motif, called motif T, is not found in RTs or other proteins, suggesting that it may be specific to the telomerase family.   

Structure of RNA splice variants

In addition to the cDNA sequence encoding the 1132 amino acid telomerase (referred to as “reference” or “wild-type” telomerase), cDNAs were found that encoding differing lengths of telomerase.  In fact, the initial cDNA clones characterized by Nakamura et al. and Meyerson et al. (Nakamura et al. 1997; Meyerson et al. 1997a) had a 182 bp deletion that resulted in a shortened telomerase due to presence of a premature stop codon. 

As it turns out, transcription of the single-copy gene produces a number of variant telomerase transcripts.  While assaying for expression of telomerase by amplification of cDNA synthesized from normal cells, immortalized cells and tumor cells RNA, Kilian et al. noticed multiple amplification products (Kilian et al. 1997).  The amplification products were isolated, and their DNA sequence determined.  Some of the products contained insertion of sequence and others deletion of sequence relative to the reference sequence.  One of the deletions (β – 182 bp) corresponded to the deletion found by Nakamura et al. and Meyerson et al. 

Fig. 3.  Alternative sequences used in splicing of hTERT transcript.

Based on cDNA analysis, at least six different alternative sequences appear to be retained in mRNAs (see Figure 3, which shows 5 of the 6 alternatives).  The sixth and 5'-most alternative sequence (not shown in Fig. 3) has an unknown length.  The nucleotides that are inserted or deleted relative to the reference telomerase are called “alternative sequences”.  Because these sequences are derived from partial or entire exons and introns, “alternative sequences” is chosen as a neutral term. 

The human telomerase gene is composed of 16 exons and 15 introns spanning approximately 40 kb of chromosome 5 (Wick et al. 1999; Leem et al. 2002).  The size of the exons ranges from 62 to 1354 bp in length.  The reference hTERT mRNA (Fig. 4) contains all 16 exons.

Fig. 4.  Reference hTERT mRNA structure.

For each of the known RNA splice variants, the effect of the presence/absence and location of each alternative sequence is presented on the assumption that it is the only alteration.  It will be appreciated that a particular alternative sequence may alter the sequence of the translated product, regardless of whether other alternative sequences are spliced in or out.  For example, the presence of alternative sequence 1 results in a frameshift and truncated protein, regardless of whether alternative sequences a, b, 2 or 3 are spliced in or out. 

Alternative sequence X is derived from intron 3.  Alternative sequence X was found inserted between bp 1769 / 1770, but was of unknown length (Kilian and Bowtell - WO 99/01560).  Intron 3 is 2089 bp, thus sequence X could be that long.  Because of stop codons present in all three reading frames of alternative sequence X, hTERT that contains X would result in a truncated protein that contains approximately 600 N-terminal amino acids and lacks all of the RTase motifs. 

Alternative sequence 1 (Kilian and Bowtell - WO 99/01560), inserted at nucleotide 1950/1951, contains the first 38 bp of intron 4, which is 687 bp long.  Its presence in mRNA causes a frame-shift and ultimate translation of a truncated protein due to a stop codon at nt 1973.  This truncated protein contains only RTase domains 1 and 2. 

Alternative sequence a (Kilian and Bowtell - WO 99/01560), located at bp 2131-2166 in the reference hTERT, is frequently observed spliced out of telomerase mRNA.  Because these 36 bp are only part of exon 6, presumably the in-frame deletion of this variant results from an alternative 3’-splice acceptor sequence in exon 6.  A protein translated from such an RNA is deleted for 12 amino acids, removing nearly all of RTase motif A.  This motif appears to be critical for RT function; a single amino acid mutation within this domain in the yeast EST2 protein results in a protein that functions as a dominant negative and results in cellular senescence and telomere shortening. 

Another RNA splice variant is deleted for alternative sequence β (Kilian and Bowtell - WO 99/01560).  The deletion encompasses all of exons 7 and 8 -- bp 2286-2468 – and encodes a truncated protein, due to a reading frameshift at bp 2287 leading to a termination codon at bp 2605.  This variant protein has RTase domains 1, 2, A, B, and part of C, but lacks a motif (AVRIRGKS SEQ. ID NO:90)  identified in the b sequence.  The motif matches a P-loop motif consensus AXXXXGK(S) found in a large number of protein families including kinases, bacterial dnaA, recA, recF, mutS and ATP-binding helicases (Saraste et al. 1990).  The importance of the P-loop in hTERT remains to be investigated.    

Alternative sequence 2 – unknown length, derived from intron 11, which is 3801 bp – is inserted between bp 2843 / 2844 (Kilian and Bowtell - WO 99/01560).  The sequence contains an in-frame termination codon at its extreme 5-end, resulting in a truncated protein of 948 amino acids, which has the entire RTase domain region, but lacks the C-terminus.  Mutations constructed in the C-terminus region (from about aa 926 to 1132) revealed that regions E-I to E-IV (located from 926-1100) are essential for catalytic and biologic function of hTERT, while mutations at the extreme C-terminus (aa 1127) generated a catalytically active but functionally dead protein (Banik et al. 2002).  Cells expressing the +1127 mutant failed to immortalize due to shortened telomeres.  Whatever the exact mechanism that causes this phenotype, it is clear that the C-terminus plays a critical regulatory role in humans, and that the product of the RNA splice variant containing alternative sequence 2 lacks the C-terminus. 

In addition to the variant above, the product of a second splice variant also lacks the reference C-terminal domain (Kilian and Bowtell - WO 99/01560).  In this splice variant, alternative sequence 3 (derived from the 3’-most 159 bp of the 781 bp intron 14) is inserted at bp 2157/2158.  An in-frame stop codon within alternative sequence 3 results in a protein having an alternative C-terminal domain.  Furthermore, the coding region donated by alternative sequence 3 contains a potential SH3 binding site, SGQPEMEPPRRPSGCVG, which matches the consensus c-Abl SH3 binding peptide (PXXXXPXXP) found in proteins such as ataxia telangiectasia mutated (ATM).  Curiously, this motif is also found near the N-terminus of hTERT (HAGPPSTSRPPRPWDTP, amino acids 304-320). 

A transcript lacking all of exon 11 (189 bp spanning nucleotides 2655 to 2843) was found in hepatocellular carcinoma cell lines (Hisatomi et al. 2003).  The protein encoded by this variant doesn’t have a reading frameshift and is 63 amino acids shorter than the reference protein.  Exon 11 contains RT motifs D and E, suggesting that the splice variant is missing residues from the catalytic site of the protein.   

Finally, an RNA splice variant was found to contain 600 bp of intron 14 fused directly to exon 16 (Wick et al. 1999).  An in-frame stop codon close to the 5’-end of intron 14 causes a prematurely terminated hTERT protein.

Expression of RNA splice variants

The critical issue is the importance and specific role of hTERT splicing variants in regulation of telomerase activity and as a potential marker of health status and survival.  Despite the provocative predictions for activities of products of splice variants, little direct evidence has accumulated.   

Of particular interest however, the splice variant deleted for alternative sequence a is a dominant-negative inhibitor of hTERT activity in cell lines (Colgin et al. 2000; Saraste et al. 1990; Yi et al. 2000).  The variant hTERT that causes these effects lacks a mere 12 amino acids, although these 12 aa contain conserved RT motif A.  When over expressed in immortalized fibroblasts and carcinoma cells, the variant inhibited telomerase activity.  Furthermore, these cells also exhibited progressively shortening telomeres and eventually apoptotic cell death or a senescence-like state.  Importantly, these data suggest that telomerase activity is controlled in part by post-transcriptional events. 

Most of the data regarding telomerase splice variants is descriptive, mainly observations of the quality and quantity of splicing variants of hTERT in normal, immortal, and cancer cells.  Although some correlations are observed (e.g., in kidney, different splice variants appear dependent on the level of telomerase activity (Ulaner et al. 1998) a number of different variants are expressed in cancer cells (Barclay et al. 2005; Fujiwara-Akita et al. 2005; Fujiwara et al. 2004; Hisatomi et al. 2003; Nagao et al. 2004; Ulaner et al. 2000; Yokoyama et al. 2001) without apparent commonality.  The complexity of the system and low expression and activity levels in some cells add to the difficultness of deciphering the intricacies of telomerase regulation. 

A word or two about nomenclature

“Telomerase” is used in this landscape paper to mean the catalytic subunit (i.e., protein with enzymatic activity) of the telomerase complex.  Some use the term “telomerase” to mean the whole ribonucleoprotein complex, of which the enzyme is but one component.  Thus, when reading articles or patents about telomerase, the reader is cautioned to confirm how the term is used in the context of the document. 

Furthermore, in early papers and patent applications, the human protein and gene name of the catalytic subunit was variously designated as hTRT, hEST2, and hTCS1.  While hTRT is still sometimes seen, the consensus term is hTERT.  In this report, hTERT is used regardless of the nomenclature in the original document. 

Methodology & Guidelines

The purpose of this patent landscape analysis is to inform the reader of the major patents and players in the area of telomerase.  It is intended as a source of scientific information or to facilitate risk management by uncovering patented technology that may need to be licensed or avoided in order to make and use a product.

Strategy for finding patent documents

Searches were performed using keywords (e.g., telomerase, splice variants), company names (e.g., Geron, Bayer), and inventors’ names (e.g., Shay).  Both Patent Lens and PatBase (www.patbase.com) database were queried.  The general approach to finding the most relevant patent documents is a scan of titles, abstracts, and lead claim when available.  Generally, scanning titles for relevance is the quickest and easiest way to comb through many documents; when a title or abstract was suggestive, but inconclusive about the claimed subject matter, the lead claim was reviewed.  In this way, a manageable list of about 100 documents was obtained.  A second review of this list entailed a review of the independent claims.  Ultimately, a final list of documents was selected. 

Decisions were made to exclude certain documents.  In particular, if the only document in the patent family were a Japanese patent or patent application, the document was omitted from analysis.  The only substantive information from these documents was a translation of the abstract, which may not accurately reflect the claims.  In addition, for areas that contained U.S. patents, U.S. applications were not analyzed.  Updates of this report will consider any that have been granted in the interim. 

Limitations of patent searches

Two main types of limitations may affect the outcome of a search: those that are inherent in the data and those resulting from the search process.  While every effort is made to minimize the limitations and their effects, inevitably there will be some.   These limitations and their effect are discussed in this section to promote appropriate and informed reliance on the patent data and analyses. 

Data are not error free.  Common types of data problems include misspellings, alternate spellings especially of names and assignees,  inconsistency of patent examiners in classifying patent subject matter,  translation errors (especially for Japanese documents, but also for European patent documents).  Some of these errors can be overcome by using wildcards in formulating the search or broadening the scope of the search; however, other errors inherent in the data cannot be neutralized.

Another source of possible problems arises from the INPADOC data, which includes patent information of more than 70 countries.  Because countries use a variety of different update schedules (two weeks to one year) and report different types of information (e.g., not all report legal status), it is not always possible to verify filing or legal status of a patent application in national patent offices. 

Furthermore, the search process may engender additional consequences.  Although using a combination of key term and classification criteria will usually capture all the key patents and applications in a field, the downside is that too many documents to realistically further screen may be found.  When there is a substantial number of documents recovered (more than a few hundred), or the primary document is not in English, the only practical means of identifying potential key documents is on the basis of the title and abstract, which may or may not represent an accurate picture of the claimed subject matter. 

As well, patents are dynamic.  On a steady basis, new patent applications are filed and published, new patents are granted, patents are abandoned, and the law changes too.  Therefore, this report is but a snapshot of the landscape as it appeared in September 2006. 

Claim construction rules

The interpretation of a claim is primarily based on the plain language of the claim and on definitions, explicit or implicit, in the text of the patent document.  Of course, claims in patent applications have not been examined and may well be different when granted (if granted at all).  For granted claims, a more precise interpretation requires a reading of the prosecution history – the written record of the examination process between a patent office and the patent owner – for instances of the patent owner stating what the claims mean or giving up claim scope. 

Patent claims are intended to provide notice to others of the scope of the protected invention.  Unfortunately, more often than not, reality falls short of the ideal.  Part of the challenge, when reading a claim, is to set aside your own biases about what words mean and question the definition of each term.  While technical terms beg definition and thus often become the subject of dispute, common words may also be contested.  Recently, patent claim construction in litigation has sometimes centered on disputes regarding definitions of words as “about” and “adjacent”, among others. 

Claims often use terms, especially technical terms, that require a bit of detective work to understand their meaning.  The steps of investigation include identifying terms that need defining (generally, words other than common English words), looking in the text of the patent document for an explicit definition or use of the term, look in the prosecution history for definition or use of the term, and consulting a dictionary or expert in the field, if available.  Pursuit of some of these steps is not within the scope of this report; the order that the steps are performed and the weight given to each is currently the subject of hotly contested litigation, but for our purposes starting with the plain language of the claim and incorporating definitions from the text of the patent will provide a useful working interpretation. 

While reading claims, a very important rule of claim construction is useful to keep in mind:  when the term “comprising” is used, a device that has more elements than the claimed device falls within the scope of the claim.  Thus, if the claim recites “an antibody comprising A, B, and C” and there is an antibody made up of A, B, C, and D, the second antibody falls within the claim.  Conversely, an antibody made up of A and B does not fall within the claim. 

Other useful rules: 

• the term “a” or “an” means one or more;
• a patentee can write her own definitions of terms;
• claims are interpreted by the current law, not the law at the time of grant.

Note that in addition to issued patents, patent applications are included in this report.  Sometimes, the claims of patent applications are written so broadly that it is not readily apparent what claim language is likely to be granted, if there is a grant. 

In the following section of this report, an overview of the patent landscape is presented.  Attention is drawn to the key patents – those that appear to have the broadest claims.  For a number of reasons, the report analyses mainly patents sought or granted in the United States:  the largest key player, Geron, is based in the United States, U.S. patents are readily obtained, the law regarding claim interpretation is most well developed.  In contrast, because patents granted by the European Patent Office are interpreted according to each country’s laws, the body of patent law isn’t homogeneous. 

The documents discussed in this report are a mix of patent and patent applications.  For patent applications, caution is warranted when evaluating the claim scope because the claims have not been examined and may differ if and when granted. 

The format of the survey generally contains a summary of the patent description, a listing of the most relevant independent claim(s) and discussion of the claim meaning. 

Disclaimer

The analyses presented herein are informational ONLY.  They are not to be construed as specific legal advice.  As advice on infringement depends upon the specific circumstances of each party and the laws of the country where the allegedly infringing activities take place, nothing provided herein should be used as a substitute for advice of counsel on the particular matter at issue.

Patent landscape of telomerase

The patent documents discussed in this report are divided into four categories depending on the protected subject matter:

• Wild-type telomerase
• Variants of telomerase
• Diagnostics
• Stem cells

The selected patents and patent applications were chosen as those with the broadest claims in the area. 

Wild-type telomerase

This section presents patents that claim aspects of nucleic acid sequences encoding wild-type^[1] telomerase.  What is “wild-type” telomerase?  For this landscape analysis it is defined as the catalytic protein containing 1132 amino acids in human and homologues in other species.  This definition, prevalent in scientific literature and in patents, has been arrived at somewhat arbitrarily.  It came about from the first publications describing the cloning of human telomerase (Meyerson et al. 1997e; Nakamura et al. 1997) in which two sequences of telomerase were found, but only one of which had an open reading frame encoding the known motifs of telomerase. 

[1] In some publications, the term “reference telomerase” is used synonymously with “wild-type telomerase”.

Human gene sequences

Not surprisingly, the co-assignees Geron Corp. and University of Colorado (current assignee of University Technology Corp., the former Colorado University organization that handled licensing for all CU campuses) dominate this area, in sheer numbers of patents as well as in scope.  A minimum of 16 U.S. patent applications claim priority from an initial filing in October 1996 (which did not contain any human-derived nucleotide sequences).  At least 10 of these applications have resulted in granted U.S. patents, some of which are directed to human telomerase and others to homologues from other species.  In addition, numerous patent applications (up to about 55) were filed in countries other than the United States.  Many of these have been granted.  

Before launching into a discussion of the Geron / University of Colorado patents, a few comments are warranted on the likely disposition of a patent application filed by Whitehead Institute (Boston, MA).  A series of U.S. provisional patent applications were filed between February and October 1997, which were based on the work of the Robert Weinberg lab (Meyerson et al. 1997e).  A PCT application (WO 98/37181) was also filed, but never converted into national applications in the non-U.S. designated countries.  Because U.S. patent procedures were conducted in secret until 2000, nothing is known about the status of any U.S. patent applications.  Because no patent has issued to date, any U.S. applications most likely have either been abandoned or are in interference^[1] with one or more Geron / UC patents. 

Fig 5.  Brief history of Geron / University of Colorado selected patent application filings

Following a series of rapid succession of filings of patent applications on 18 April, 25 April, and 6 May 1997, containing incomplete sequences of human telomerase, on 9 May 1997, the Geron / University of Colorado groups filed a patent application with the full-length sequence.  US 6,261,836 B1 was granted on 17 July 2001 from this filing; its claims are directed to “human telomerase reverse transcriptase protein”. 

With respect only to the disclosure related to human telomerase, the application details the cloning of hTERT.  Briefly, partial homologous sequences were identified in a BLAST search of publicly available EST sequences queried against the Euplotes and Schizosaccharomyces telomerase sequences.  A match was scored for an EST sequence derived from a partial cDNA clone (GenBank accession AA281296).  When the open reading frame of this clone was aligned to the query sequences, the presence of signature motifs strongly suggested that this was the long-sought after human telomerase.  Using standard techniques, additional 5’ sequence information was obtained as well as a lambda vector clone (l25-1.1) that had complementary sequence that was sub-cloned into a plasmid (pGRN121 – ATCC accession no. 209016).  The sequence of the insert revealed the entire open reading frame encoding the human telomerase protein.  Interestingly, pGRN121 contained an insertion of 182 nucleotides relative to the GenBank accession AA281296.  The open reading frame of pGRN121 encoded a protein of 1132 amino acids having a molecular weight of approximately 127,000 daltons.  SEQ ID No: 224 presents the nucleotide and amino acid sequences of hTERT.

The independent claims are directed to nucleic acid sequences encoding the hTERT protein, hTERT protein encoded by the sequences, cells comprising the nucleic acid sequence, and a method of preparing the telomerase complex using the claimed hTERT protein and a telomerase RNA component. 

Claim 1 asserts a “synthetic or recombinant human telomerase reverse transcriptase” (hTERT)

• protein;
• variant of the protein;
   o        variant is encoded by a polynucleotide that hybridizes under stringent conditions to a complement of SEQ ID NO: 224 (hTERT); or
• fragment of the protein;

AND

• the protein, variant or fragment has telomerase catalytic activity when complexed with a telomerase RNA component.

Literally read, the claim terms, “hTERT protein”, “variant of the protein”, and “fragment of the protein” extend beyond the disclosed amino acid sequence.  The term “hTERT protein” has no limitation of specific sequence, as would have been expected.  Its only limitation is the requirement of catalytic activity (discussed below).  But what exactly falls within the scope of “hTERT protein”? 

Reading the claim literally, any human TERT protein that has catalytic activity would be encompassed.  Any hTERT protein could arguably include polymorphisms, RNA splice variants, deletions, etc.  It’s unclear whether or not the modifier “human” means that the sequence has to be a natural sequence.  If the telomerase is not a natural sequence found in humans, then what makes a telomerase a “human” telomerase?  Another reason to believe that only natural sequences, e.g., the amino acid sequence of SEQ ID NO: 225 and polymorphisms, are encompassed is the recitation of “variant of the protein” as an alternative element.  If the term “protein” also included non-natural sequences such as those encoded by hybridizing polynucleotides, then there wouldn’t have been any need to recite “variants” in the claim. 

Reading the claim in light of the specification, however, an argument can be made that the term only encompasses the specific amino acid sequence set forth in SEQ ID NO: 225, because that is the only amino acid sequence shown.  (The Federal Circuit Court applies a strict written description requirement to biotechnology inventions, which usually limits the claims to the precise sequences disclosed in the patent application.)   This interpretation however presumes that a court would find it so.  As it stands in 2006, though, most patent lawyers lament the lack of certainty of claim interpretation.  

While by no means certain, the term hTERT protein probably means just SEQ ID NO: 225.  This conclusion is supported by the presumed meaning of “fragment thereof”, which refers to sequences found within SEQ ID NO: 225 – see below for more discussion.  If hTERT refers just to SEQ ID NO: 225, then polymorphisms of hTERT and RNA splice variants are excluded. 

“Fragments” of hTERT are contemplated to be “of various lengths. In one embodiment, the portion of polypeptide comprises fragments of lengths greater than 10 amino acids. However, the present invention also contemplates polypeptide sequences of various lengths, the sequences of which are of which are included within SEQ ID NOS: …225, from 5 to 1100 amino acids (as appropriate, based on the length of (SEQ ID NOS: ….. 225).”  So, from this description, is the minimal length of a fragment 5 amino acids?  10 amino acids?  Or some other length?  Because of the requirement that the fragment have catalytic activity, the minimum length of a fragment is not terribly relevant.  To have catalytic activity, a fragment of hTERT would need to be substantially longer than 5 or 10 amino acids.  What we can infer from the description however is that the fragment sequence is found within SEQ ID NO: 225.  So, fragments of polymorphic hTERT or RNA splice variant hTERT do not appear to be encompassed by this element of the claim. 

On the other hand, “variant of the protein” carries a definition in the claim itself – a protein encoded by a hybridizing nucleic acid, which hybridizes under “stringent conditions”.  According to the patent application, “stringency" typically occurs in a range from about 5° C to about 20° C to 25° C below Tm of the probe.  From the point of view of hybridization kinetics, this is not much of a limitation.  While it is often stated that the maximum divergence of sequences that hybridize at about 20-25° C below Tm is 25% mismatch, it is the stringency of the wash conditions that generally controls the amount of mismatch (or conversely, the amount of identity) observed.^[2]  The common hybridization condition of 20-25° C below Tm is a consequence of experiments begun in the 1970s by Wetmur and Davidson, who determined that this temperature resulted in maximal hybridization efficiency, not necessarily maximal hybridization specificity.  Even though it is an erroneous belief, many, or most, scientists would consider the stated stringent hybridization conditions to yield nucleic acid molecules having at least 75-80% identity to the probing sequence.  The various RNA splice variants would hybridize under these conditions, as would many related sequences.  Because of the activity requirement, at least some of the splice variants would necessarily be excluded from the claimed telomerase molecules. 

Any protein / variant / fragment of hTERT has a major limitation in that it must have “telomerase catalytic activity” when complexed with a telomerase RNA.  While this term is not explicitly defined, the most probable meaning is that the catalytic activity is using a “portion of its internal RNA moiety as a template for telomere repeat DNA synthesis” and more specifically, “extend[ing] the G strand of telomeric DNA.”  (col. 3, lines 17-30) Example 4 provides the method for assaying telomerase activity; no other assay appears to be described.  The assay is for extension of dGTP on an oligonucleotide substrate; assay conditions do not mention the inclusion of a telomerase RNA component, however. 

US 6,921,664 has identical disclosure to US ‘836.  The claims however recite a recombinant expression vector comprising an encoding region for telomerase protein, a variant or a fragment.  As for the claims of US ‘836, the protein, variant or fragment must have telomerase catalytic activity when complexed with a telomerase RNA.  Similarly, the nucleic acid encoding telomerase hybridizes to the complement of SEQ ID NO: 224 under stringent conditions. 

Because of how the claim is written, it appears that a nucleic acid sequence that incorporates enough codon differences (taking advantage of codon degeneracy) wouldn’t fall within the scope of this claim because it couldn’t hybridize to the complement of SEQ ID NO: 224.

US 6,927,285 has a disclosure written prior to establishing full-length telomerase sequence.  Only a partial sequence, replete with sequence errors, was presented.  To overcome that deficiency, Geron / University of Colorado made a biologic deposit of the plasmid pGRN121, which contained an insert encoding the full-length telomerase protein.  Under U.S. patent law a biological deposit can satisfy the enablement - and written description -  requirements.   The deposit was therefore a wise decision. 

This patent has six independent claims to the telomerase sequence.  With reference to pGRN121, both claim 1 and claim 5 are directed to an isolated cDNA encoding hTERT, wherein the cDNA is contained in pGRN121 (claim 1) or wherein the cDNA hybridizes to the insert of pGRN121 (claim 5). The cDNA of Claim 2 has the restriction map of pGRN121. 

Claims 3 and 6 claim an isolated nucleic acid encoding a “naturally occurring” hTERT or variant thereof (claim 3) and an isolated cDNA encoding hTERT (claim 6), wherein the nucleic acid or the cDNA hybridize to the incomplete sequence presented as SEQ ID NO: 173.  Claim 4 recites SEQ ID NO:173 with a 5’-Met codon. 

Usually claims to a deposited sequence would be considered to be relatively narrow.  Essentially such claims are equivalent to a claim to SEQ ID NO: N.  In a twist however, claims in this patent recite hybridizing sequences, effectively broadening the scope.  The requirement exists though that the nucleic acids must encode human telomerase reverse transcriptase protein.  As discussed for other patents in this section, a plausible interpretation is that the claim encompasses the “wild-type” or full-length telomerase and possibly polymorphisms. 

A continuation-in-part of US 6,261,836, the claims of US 6,475,789 are directed to mammalian cells that contain a recombinant nucleic acid sequence that encodes hTERT.  The disclosure of US ‘789 first describes the consensus motifs that are characteristic of TERT proteins and nucleic acids identified in species such as Oxytrichia, Schizosaccharomyces, Euplotes, and human.  Most of the remaining disclosure provides pages of detail about the human gene and protein, including how to clone the gene, recombinant expression of the protein, purification of the protein, and raising antibodies to the protein.  Multiple activities of the protein are discussed along with descriptions of assays for the activities.  The activities include reverse transcriptase activity, telomere binding, dNTP binding, and telomerase RNA binding. 

Uses for the nucleic acids and proteins comprise the remainder of the disclosure.  The uses include treatment of cancer and other diseases and conditions, vaccine production, and increasing the proliferative capacity and production of immortalized cells and animals.  As well, diagnostic assays for the presence of telomerase are provided.  These assays are proposed for diagnosis and prognosis of cancer and other conditions and monitoring cells in culture, among other uses.

Although the disclosure is hefty, because of U.S. patent law, claims are limited to one invention – as determined by the U.S. Patent Office.^[3]  In this case, the eight claims are directed to:

A mammalian cell that contains

·         a nucleic acid sequence that encodes TERT protein, variant or fragment having telomerase catalytic activity;

o        the TERT sequence hybridizes to the hTERT sequence (ID NO: 1) using stringent conditions. 

The terms and hybridization conditions have all been discussed above for US ‘836.  Whereas the claims of US ‘836 claim a TERT protein (or variant or fragment), the claims here are directed to nucleic acid sequences.  Because of how the claim is written, it appears that a nucleic acid sequence that incorporates enough codon differences (taking advantage of codon degeneracy) wouldn’t fall within the scope of this claim because it couldn’t hybridize to SEQ ID NO: 1. 

The title of US 6,444,650 is “Antisense compositions for detecting and inhibiting telomerase reverse transcriptase.”  This very slim patent is directed to antisense oligonucleotides that specifically anneal to hTERT nucleic acid molecules.  These include not only sequences encoding hTERT, but also any upstream, flanking, noncoding, and transcriptional control elements, hTERT pre-mRNA, mRNA, cDNA, and the like.  Antisense is considered to be at least 7 nt up to about 100 nt, but is often in the range from 10-50 nt.  The inventors prefer approximately 30 nt long antisense. 

Like their other patent applications, “specific binding” or “specific hybridization” is that annealing to a target polynucleotide that occurs under stringent conditions.  In this patent, “stringent conditions” are also defined as from about 5 to about 20 to 25° C below Tm of the target sequence in 1 M NaCl.  (See discussion about this definition under US 6,261,836.)   Other definitions of terms are broader in this disclosure than in earlier disclosures – e.g., hTERT activity now refers to one or more of activities found in naturally-occurring full-length hTERT protein. 

The lead claim recites antisense oligonucleotides that

·         hybridize to SEQ ID NO: 1 under stringent conditions; and

·         inhibits expression of hTERT mRNA. 

Although the text of the patent discusses antisense oligonucleotides that anneal to other than hTERT coding sequences, the claim limits them to annealing to SEQ ID NO: 1.  This sequence contains approximately 50 bases of 5’-noncoding sequence, the entire hTERT coding sequence, and approximately 575 bases of 3-noncoding sequence.  According to the claim, the antisense could have a different complementary sequence than found in SEQ ID NO: 1 as long as the antisense anneals under the stated conditions.  As is well known, hybridization kinetics and duplex stability is not the same for oligonucleotides as it is for polynucleotides over about 600 bases long.  Base mismatches in short duplexes can significantly affect thermal stability.  With these considerations, this claim is unlikely to encompass widely divergent oligonucleotide sequences; a prediction of the likely boundaries is outside the scope of this landscape. 

Rounding out the group of Geron patents, US 6,610,839 claims hTERT promoter sequences.  A human genomic DNA library was screened to identify a clone containing hTERT coding sequences.  One isolated clone (lGF5) contained approximately 13 kb of DNA upstream of the start site of the cDNA sequence.  Subfragments of this region were tested for promoter activity by linking them to a reporter gene sequence.  Claims were granted to the deposited lambda phage clone, hybridizing nucleic acid sequences, and specific sequences, along with 80% identical sequences. 

** Patents and applications that are published in English and that have claims similar to the listed United States patent are specifically mentioned. The claims needed to be quite similar before it was included.  Even if not mentioned, a non-U.S. patent could have claims that encompass similar subject matter.  For example, a patent claiming a cell having an introduced nucleic acid encoding telomerase would encompass a cell containing an expression vector expressing telomerase, but is not listed as an “equivalent”. 

Homologous telomerase gene sequences

This group of patents – three of which are owned by Geron and University of Colorado (individually or jointly), one of which is co-owned by Geron and Albert Einstein College of Medicine, and one of which is owned by Research & Development Institute in Montana – concern nucleic acid sequences and gene products of telomerase from other species.  By no means does this group exhaust the species from which telomerase gene has been cloned; telomerase from other species can be found in the scientific literature. 

The species represented in this group are:

• Candida albicans (US 6,541,202);
• Euplotes (US 6,093,809 and US 6,166,178);
• Schizosaccharomyces pombe (US 6,309,867);
• Mus musculus (US 6,767,719).

* - A terminal disclaimer was filed for this patent.  Without obtaining the full prosecution history of this patent to reveal which patent it was disclaimed in favor of, certainty of the expiry date is not possible. 

[1] An interference is a procedure conducted by the United States Patent and Trademark Office when multiple applications have identical claims.  The purpose of an interference is to determine who was the first-in-time inventor and so deserves the patent. 

[2] For a review of hybridization, see www.roche-applied-science.com/PROD_INF/MANUALS/InSitu/pdf/ISH_33-37.pdf. 

[3] Geron has many pending patent applications on telomerase, some of which may contain claims directed to disclosure that isn’t claimed in this patent. 

Concluding remarks

With only a single exception, all the patents directed to wild-type telomerase are owned at least in part by Geron Corp.  Its dominance came about by “winning the race” to clone human telomerase.  The patents discussed above include claims for telomerase protein and fragments, nucleic acids encoding telomerase, expression vectors that encode telomerase, mammalian cells containing a nucleic acid that encodes telomerase, anti-sense oligonucleotides, and promoter sequence of the gene encoding telomerase. 

Although only one telomerase sequence was disclosed, Geron’s claims encompass enzymatically active variants and fragments of human telomerase protein and nucleic acid sequences.  The range of the variant sequences is determined by hybridization conditions – conditions that are called “stringent”.  Such conditions generally result in related sequences that are at least 75% identical.  Although by no means certain, a plausible interpretation of these claims is that the variants of the claims are full-length telomerase molecules.

In addition, Geron has protected many related concepts including:  cells transfected with and expression telomerase, expression vectors encoding telomerase, anti-sense oligonucleotides, antibodies that specifically bind telomerase, and telomerase promoter sequences. 

Outside the United States, Geron has actively pursued protection in at least European countries, Australia, Brazil, Canada, China, Israel, Japan, Republic of Korea, New Zealand and Singapore.  (Geron may have filed in other countries that do not report or reliably report patent application data.)  The scope of protection varies among these countries in part due to what is protectable subject matter in each country and also to differences in application of patent laws.  In countries where there isn’t any patent protection of telomerase, it is safe for anyone to practice the invention(s) providing that the output isn’t being exported to a country where the invention is protected.  Where telomerase has been protected and you want to use one or more of the protected telomerase inventions, we suggest you obtain legal advice directed to your particular circumstance. 

Variants of telomerase

This section presents patents directed to variants (alternate forms) of wild-type human telomerase.  Variants come in many different flavors, including:  mutations, insertions, deletions, polymorphisms, and RNA splice variants.  The first set of patents presented in this section is about RNA splice variants.  Two groups initially described RNA splice variants:  CAMBIA^[1] and Bayer.  In the United States, CAMBIA obtained two granted patents, while Bayer appears to have largely abandoned their effort to obtain protection for splice variants, possibly because their disclosure of splice variants was later in time than CAMBIA’s.  Both CAMBIA and Bayer have been granted Australian patents however. 

US 6,846,662 and US 6,916,642, which are owned by CAMBIA, have identical disclosures.^[2]  The disclosure presents a number of RNA splice variants of telomerase.  Specifically, seven alternative “introns/exons” were identified, which theoretically could yield 128 different RNAs.  Because of frame-shifts or stop codons introduced by some of the alternative “introns/exons”, the number of different telomerase variant polypeptides would be much smaller.  Now that the genomic structure of the telomerase gene is known, these alternative sequences can be mapped. 

Strikingly, most of the alternative sequences do not correspond to entire exons or introns, and as such, the RNA splice variants disclosed would not have been predicted by the genomic sequence. 

The claims of the two patents are directed to proteins (US ‘662) and to nucleic acids (US ‘642).  The protein patent claims telomerase proteins having:

• specific sequences of Figure 11 (claim 1);
• proteins that are 90% identical to the specific sequences and that have either catalytic activity or binds telomerase RNA – and is not wild-type telomerase (claim 4);
• individual alternative “introns/exons” (claim 2);
• a splice variant lacking 13 amino acids from residues 711-722 (inclusive) and proteins having 90% identical amino acids as well as same requirements of claim 4 (claim 3);
• a sequence encoded by a nucleic acid having at least one of the disclosed alternative introns/exons excluding wild-type telomerase (claim 6).

The lead claim of the nucleic acid patent recites a molecule encoding a splice variant in which the variant has at least one of six alternative introns/exons, which would yield 64 different proteins.  Other claims are directed to specific DNA sequences encoding variant telomerases presented in Figure 11 and sequences encoding a protein having at least 95% identity to the variants.  Individual nucleic acid sequences of the alternative introns/exons are also claimed.  There are also claims to nucleic acid probes and amplification primers for detecting and replicating RNA splice variants.   Methods for establishing patterns of expression of the variants, e.g., for detection or prognosis of cancer, form many of the remaining claims.  While not claiming all RNA splice variants, these patents encompass nearly every splice variant identified to date. 

Australian patent AU 748442 is related to the two CAMBIA U.S. patents and has essentially identical disclosure.  The claims in AU ‘442 are broader however. 

The lead claim encompasses a “nucleic acid molecule encoding a splice variant” of wild-type telomerase.  Moreover, the wild-type telomerase reference protein is not limited to humans, but includes related telomerases that are encoded by nucleic acid molecules that hybridize under conditions of low stringency to the region containing reverse transcriptase motifs.  By limiting the hybridization region to the most conserved part of telomerase and setting the hybridization conditions to low stringency, the reference telomerases will likely include many vertebrate species.  Low stringency hybridization may include molecules with as little as 50% identity and possibly less.  Thus, the claim covers splice variants of telomerase from a variety of vertebrate species. 

In contrast to the broad claims of CAMBIA’s Australian patent, the claims in the Bayer patents AU 742489 and AU 745420 recite specific DNA sequences of splice variants.  Although the earliest priority date of AU ‘420 is before the earliest priority date of AU ‘442, splice variants weren’t disclosed in the earliest priority document for AU ‘420.  Thus, CAMBIA filed disclosure of splice variants prior to Bayer.  The lead claim in AU ‘420 recites four different variants encoded by nucleic acid sequences that are:

• deleted for 182 bp (b region);
• deleted for 36 bp (a region);
• deleted for both the 182 bp and 36 bp; or
• deleted for about 1800 bp at the 5’ end and has an alternate 3’ end.

The first three variants were described in AU ‘442 and appear to be encompassed by its claims, or they are identical claims (and probably not valid as a result).  The fourth variant is a fragment, and whether it is also covered by the claims of AU ‘442 requirse a more precise inquiry into how Australian patent law would interpret the claim of AU ‘420. 

The second Bayer patent – AU ‘489 – claims isolated intron sequences or fragments of these sequences which have a regulatory effect.  In the intron sequences, mainly in intron 2, the inventors have identified potential binding sequences for DNA-binding proteins – regulatory proteins.  The regulatory proteins include C/EBP, CRE.2, Sp1, GRE, CREB, c-Myc, CCAAT site, and Rb site.  In addition to the numerous candidate binding sites in intron 2, one potential Sp1 binding site was found in intron 1 and one potential c-Myc binding site was found in the 5’-untranslated region. 

These introns, intron 1 and 2, have not been found in RNA splice variants of telomerase to date.  Moreover, each intron sequence that is found in splice variants corresponds only to a portion of the intron.  Unless the portion specifically has a regulatory effect, and none have been identified so far, then the claims in Bayer’s patent AU ‘498 do not cover individual alternate intron/exon sequences. 

The second set of patents in this section is directed to telomerases deleted for specific regions of sequence.  All of these patents were granted to Geron.  The claims in one are directed to specific deletions that do not affect the activity of telomerase, while the claims in the other recite deletions that abolish telomerase activity.  According to the patent, a variant telomerase has “activity” if it has at least 40% of the activity of a wild-type telomerase.  “Lack of activity” means that the variant has less than 1% of “wild-type” activity; “intermediate activity” is used to describe a telomerase with between 1% and 40% activity. 

The inventors discovered that amino acid residues 192-323 or residues 415-450 can be deleted and the resulting telomerase retains catalytic activity.  Some decrease was observed of the binding of the telomerase RNA component to these variants.  More interesting however, are the variants that are deleted for one or more of residues 192-450, 637-660, 638-660, 748-766, 748-764 and 1055-1071.  These variants not only lack telomerase catalytic activity but appear to inhibit the activity of wild-type telomerase – “dominant negative variants”.  For reference, the region from residues approximately 620-902 contains the telomerase-specific and reverse transcriptase motifs, and the splice variant that is dominant negative - Da - is deleted for residues 708-722. 

The claims of US 6,337,200 are relatively straightforward.  Claim 1 recites a polynucleotide encoding an hTERT variant, which is deleted for at least 10 amino acids of residues 192-323 or 415-450.  Moreover, the variant has catalytic activity.  Other independent claims are directed to using the variants to increase the proliferative capacity of a human cell in vitro by expressing the variant in the cell and to producing the variant protein by expression in a host cell or in a cell-free expression system. 

The claims of US 7,091,021 are directed to polypeptide variants and methods of using the variants to inhibit telomerase catalytic activity.  The simplest claim (claim 3) for polypeptide variants recites a telomerase deleted for one or more of residues 326-415, 560-565, 637-660, 748-766, 930-934, 1055-1071, or 1084-1116.  An unusual aspect of this claim is the recitation that the deletions “consist[ing] essentially of” the residues already mentioned.  “Consisting essentially of” is a patent term that “limits the scope of a claim to the specified materials or steps "and those that do not materially affect the basic and novel characteristic(s)" of the claimed invention (MPEP 2111.03, 8^th ed.; emphasis in original).  It is the understanding of the author that “consisting essentially of” is rarely, if ever, used in biotechnology claims, and that the meaning of this term for biotechnology claims is uncertain. 

Claim 2 is relatively simple and is directed to a variant lacking telomerase activity, in which the variant is wild-type telomerase deleted for one or more of the following: 

• at least 10 consecutive amino acids between 326-415, 637-660, 748-766, 1055-1071, 1084-1116;
• amino acids 560-565; 930-934.

Claim 1 is similar to claim 2 except that “wild-type telomerase” is replaced by “a polypeptide encoded by DNA that hybridizes to” the complement of the sequence encoding wild-type telomerase.  Hybridization conditions are the same as for other Geron patents and have been discussed in detail elsewhere in this landscape.  An additional requirement is that the variants inhibit telomerase enzyme activity in situ.  This requirement doesn’t impose an additional limitation relative to the other claims; it is intended to exclude polypeptides that have the recited deletions but are not dominant negative variants. 

[1] The original assignees were Cambia Biosystems LLC and Peter MacCallum Cancer Research Institute.

[2] In the United States, claim sets are often subject to “restriction”, meaning that the Patent Office believes that the claims are drawn to multiple inventions.  Because only one invention can be claimed in a patent, the applicant chooses an invention for examination; the remaining inventions can be examined in separate applications.  The separate applications have identical disclosures and are called “divisional” applications. 

Concluding remarks

Are the claimed variants dominated by claims to wild-type telomerase?  The question is moot for the constructed deletion variants of the second set because they are also owned by Geron.   The claimed splice variants are not owned by Geron however, and so the answer to the question is of paramount importance.  This question can’t be definitively answered here; an answer requires additional documents and considerable legal analysis.  But, an educated guess can be hazarded. 

As discussed in the section on wild-type telomerase, some of the claims at first blush appear to read on certain RNA splice variant sequences (e.g., hybridizing molecules will encompass polymorphisms, small base changes, some insertions and deletions).  Under current United States patent law, sequences must be fully described and enabled.  Recent case law indicates that the threshold amount and type of disclosure to satisfy the written description (especially) and enablement requirements is exacting.  For the most part, claims to undisclosed sequences aren’t patentable.  Geron disclosed wild-type telomerase and did not disclose splice variants.  Moreover, Geron’s disclosures for wild-type telomerase are directed to full-length molecules.  Analysis above of the key claims for wild-type telomerase preliminarily concluded that hybridizing molecules were directed to full-length telomerase; in addition, fragments of wild-type and related telomerases aren’t likely to encompass or contemplate splice variants.  Furthermore, the key claims for wild-type telomerase require catalytic activity.  Therefore, the splice variants that do not have catalytic activity should fall outside these claims.  Taken all together, a credible interpretation can conclude that claims to splice variants of telomerase are not dominated by Geron’s claims to wild-type telomerase and related molecules. 

The impact of this conclusion, if correct, will be most keenly felt by CAMBIA^[1] and any of its licensees who will be able to practice the invention without needing permission or a license from Geron. 

[1] The author of this patent landscape has substantial connections with CAMBIA and thus is not without some bias. 

Diagnostic assays based on telomerase

Long before the gene encoding telomerase catalytic protein was isolated, telomerase was recognized as a key regulator of the replicative lifespan of cells.  It is fairly obvious that diagnostics for telomerase activity would therefore be of interest for proliferative diseases such as cancer. 

Diagnostic assays are currently the subject of some interesting patent court cases.  For example, in Innogenetics, N.V v. Abbott Laboratories, Fed. Cir. App. 2007-1145, 2007 U.S. Dist. LEXIS 3148 (W.D.Wis. 2007)(Crabb, J.), the accused infringer, the pharmaceutical company Abbott, appeals from an injunction in a challenge to recent court decisions concerning E-Bay and willingness to license.  This represents the first medical case where injunctive relief has been granted where the US court has expressly acknowledged that a patent situation might result in the patient-public being deprived of the best medical techniques on the market.  A decision is likely only in 2008-2010 because the case will not be heard until late 2007 (it is not possible to predict a timeline for disposition of a U.S. Federal Circuit appeal.

Assays for telomerase activity

As part of its drug discovery program for inhibitors, Geron scientists developed an assay for telomerase activity, which was a modification of the method of Morin (Cell 59: 521, 1989).   As disclosed in WO 93/23572, one method for determining telomerase activity is “by measuring the rate of elongation of an appropriate repetitive sequence, having 2 or more, usually 3 or more, repeats of the telomere unit sequence, TTAGGG.”  In this assay, radiolabeled nucleotides are incorporated into an elongated oligonucleotide substrate containing the TTAGGG repeats; the reaction products are resolved by gel electrophoresis and visualized.  Because the telomerase enzyme stalls and can release the substrate after adding the first G in the repeat, the pattern obtained is a six nucleotide ladder of extended substrates. 

Following on this initial assay, Geron scientists concocted a new assay, which is disclosed and claimed in US Patent 5,629,154.  As claimed, the new assay comprised:

• preparing an extract from a cell sample;
• mixing an aliquot of the extract with (i) a telomerase substrate that lacks a repeat sequence, and (ii) an oligonucleotide primer complementary to a telomeric repeat sequence; and
• detecting a double-stranded DNA reaction product, one strand of which is an extended telomerase substrate containing repeat sequences and the other strand of which is extended primer. 

The conditions for extension of the substrate and for extension of the primer are not specified in the claims and as such, the conditions are not limiting.  For example, the extension of the primer can be effected by a DNA polymerase, a template-dependent DNA ligase, or other suitable enzyme and the reaction product may be labeled by incorporation of radio-labeled dNTPs or by a 5’ label or by some other means. 

A further improvement of this basic strategy at Geron resulted in the widely used TRAP (telomeric repeat amplification protocol) assay (Kim et al. 1994).  The claimed method of US 5,863,726 comprises:

• incubating a cell sample with an exogenous telomerase substrate under conditions such that telomerase can catalyze extension of the substrate by addition of telomeric repeat sequences;
• replicating (e.g., by amplification such as PCR) the extended substrate; and
• correlating telomerase activity with presence of extended substrate and lack of activity with absence of extended substrate.

Although the claims don’t limit many of the elements, such as the nature of the substrate (e.g., lacking a telomeric repeat sequence), the method of replication, means for detecting the extended substrate, in practice most scientists perform TRAP using PCR and visualizing the amplified product by detection of radioactive-labeled nucleotides after the products are separated by electrophoresis.  As is typical of patents, dependent claims in the patent recite these specific elements as well as a variety of substitutes for the elements.   Other claims are directed to kits for detecting telomerase activity having (i) a telomerase substrate, and (ii) a primer that is complementary to a telomeric repeat sequence. 

Further consolidating Geron’s position on assays for telomerase activity, US 5,804,380 claims using the TRAP assay for screening candidate modulators of telomerase activity (claim 1).  In addition, Geron claims a kit for detecting telomerase activity similar to the one in US ‘726 except that the substrate is labeled and instructions are included.  This kit claim actually falls within the scope of the claim in US ‘726, but is directed to what is probably a more likely format for commercial kits. 

Geron has not left a lot of room for others in this area, as indicated by the dearth of other pending or issued patents.  CTRC Research Foundation in Texas however has a patent, US 5,856,096, that claims a ligation-style assay for telomerase activity.  In their method, activity is detected in a sample by:

• incubating a sample with a telomerase substrate and dNTPs to form an extended substrate; then
• adding two oligonucleotides to the extended substrate, in which the two oligonucleotides are adjacent to each other when annealed to the extended substrate;
• adding a ligase that will ligate together the two oligonucleotides; and
• detecting the ligated product. 

Other claims are directed to use of assay to screen inhibitors of telomerase activity and to distinguish between processive and nonprocessive activities of telomerase.