RICH, Circuit Judge.

I. Background

Genentech, Inc. (Genentech) appeals from the grant, by the United States District Court for the Northern District of California in No. C-94-3334 CW, 1995 WL 450846 (N.D. Cal. 1995), of summary judgment in favor of Chiron Corporation (Chiron) in an interference between Genentech and Chiron involving a DNA construct on the ground that Genentech's claimed invention is not within the scope of the sole interference count as properly interpreted. We reverse and remand to the district court for further proceedings in accordance with this opinion.

II. Discussion

This case arises from Patent Interference No. 102,208 in the United States Patent and Trademark Office (PTO) between two copending applications: United States Patent Application No. 06/506,078, entitled "Preparation of Human IGF and EGF Via Recombinant DNA Technology," (Lee application) by inventors James M. Lee, Axel Ullrich, and Arjun Singh (collectively Lee) and United States Patent Application No. 06/922,199, entitled "Hybrid DNA Synthesis of Mature Insulin-Like Growth Factors," (Barr application) by inventors Philip J. Barr, James P. Merryweather, Guy Mullenbach, and Mickey S. Urdea (collectively Barr). The Lee application was filed on June 20, 1983 and the Barr application was filed on October 23, 1986 and accorded the benefit of United States Application 06/487,950, which was filed on April 25, 1983. Therefore, Lee, whose application was assigned to Genentech, was the junior party and Barr, whose application was assigned to Chiron, was the senior party.

The sole count of the interference is identical to claim 22 of Barr's application, and reads:

A DNA construct comprising a sequence coding for human insulin-like growth factor-I joined in proper reading frame with Saccharomyces alpha-factor secretory leader and processing signal sequence.

Saccharomyces, commonly known as baker's yeast, naturally secretes a small, thirteen amino acid pheromone known as alpha factor during its reproductive cycle. As a first step in this process, the yeast cell produces a precursor protein from the DNA sequence encoding alpha factor that consists of a secretory leader, which serves to signal the cell to process the precursor protein, followed by four copies of the alpha-factor protein separated by processing sequences of six or eight amino acids each. Each processing sequence is recognized and excised by an enzyme in the yeast cell to release the four copies of the alpha factor outside of the yeast cell. As a result of this process, four copies of mature alpha-factor protein are secreted from the yeast cell for each precursor protein.

Human insulin-like growth factor-I (IGF-I), is a growth-promoting protein that mediates the effect of human growth hormone. It is undisputed that human IGF-I consists of a specific sequence of seventy amino acids that was published in 1978. See Ernst Rinderknecht and Ren Humbel, The Amino Acid Sequence of Human Insulin-like Growth Factor I and Its Structural Homology with Proinsulin, 253 J. Biol. Chem. 2769 (1978).

Lee and Barr sought to produce mature IGF-I by adapting the Saccharomyces DNA encoding the gene for the precursor protein. Utilizing a DNA construct with only one copy of alpha-factor protein, both Lee and Barr replaced the DNA sequence of the precursor protein encoding alpha factor with a DNA sequence encoding IGF-I. This new DNA construct is then inserted into an expression plasmid, and transformed into a yeast cell. The goal of both Lee and Barr was for such a transformed yeast cell to secrete some form of human IGF-I. In replacing the DNA sequence encoding alpha factor with that encoding IGF-I, Lee, however, added twenty-seven additional nucleotide bases constituting a collagenase cleavage site between the DNA coding for the alpha-factor processing sequences and human IGF-I. When this construct is inserted into a yeast expression plasmid and transformed into a yeast cell, the cell secretes a "fusion protein" or "modified IGF-I" consisting of a collagenase cleavage site at the carboxy terminal of human IGF-I.

A.

In analyzing Lee's case for priority, the Board of Patent Appeals and Interferences (Board) rejected Barr's argument that Lee's DNA construct fell outside of the scope of the count because when inserted into a plasmid and transformed into a yeast cell, it would cause the yeast cell to produce a fusion protein. The Board held:

The interfering subject matter, as defined by count 1, is directed to a DNA construct comprising two components, i.e., a DNA sequence coding for human IGF-I and a Saccharomyces alpha-factor secretory leader and processing signal sequence. Each of these components are [sic] made up of nucleotides and must be present in Lee's DNA construct in order to fall within the scope of the count. As long as these two components are present, other materials may also be present because the term "comprising" permits their inclusion. Thus, we view count 1 as being directed to a DNA construct which encodes IGF-I as either a mature protein or a fusion protein. Since the DNA construct made by Lee encodes IGF-I as a fusion protein, the Lee proofs fall within the scope of the count.

Lee v. Barr, Patent Interference No. 102,208, Slip. Op. at 8 (BPAI July 19, 1994).

The Board held, however, that Lee had failed to prove any practical, therapeutic utility of its fusion protein including a collagenase cleavage site in addition to IGF-I. Therefore, the Board awarded priority to the senior party, Barr.

B.

Genentech, the assignee of the junior party Lee, filed a civil action under 35 U.S.C. § 146 in the United States District Court for the Northern District of California to challenge the Board's award of priority to Barr. Chiron, the assignee of the senior party Barr, moved for summary judgment, again alleging, as it had before the Board that Genentech's claimed invention of a DNA construct encoding modified IGF-I was not within the scope of the interference count as properly interpreted. The district court granted Chiron's motion for two reasons.

First, the district court found that the count was directed to a DNA construct that codes for authentic or mature, human IGF-I. The district court found that the term "coding" in the field of biology connotes the protein that results from expression of a particular DNA. Therefore, the court concluded that "[s]ince Genentech's DNA construct codes for a sequence of 79 amino acids, consisting of human IGF-I plus nine additional amino acids, rather than for human IGF-I alone, it is outside the plain meaning of the language of the count." Genentech Inc. v. Chiron Corp., 1995 WL 450846, at *4 (N.D. Cal. 1995).

Second, the district court agreed with Chiron that the only interpretation of the word "joined" required that the two elements to be joined must be directly connected with no intervening material. Id. at *5. The district court explained that this interpretation does no violence to the open ended nature of the claim resulting from the use of the term "comprising" or the phrase "in proper reading frame." The district court interpreted "comprising" to permit additional material before and after the two required elements, but not in between the two elements that are joined. Id. The district court found that the phrase "in proper reading frame" was not rendered superfluous by requiring the elements to be joined directly because the second element when directly connected is not necessarily in proper reading frame with the first element. For instance, if the second element were connected in reverse orientation or the two elements had overlapping codons, the district court asserts that the two elements would be directly connected, but not in proper reading frame. Id. Based on this reasoning, the district court affirmed the award of priority to Chiron without reaching the utility issue. Genentech appeals this decision of the district court. We reverse.

C.

Genentech asserts on appeal that the district court gave an improperly narrow interpretation to the count; the district court improperly determined a factual issue in dispute, i.e. whether, under 35 U.S.C. § 112, first paragraph, Chiron had enabled the invention of the count as defined by the district court requiring secretion from the yeast cell of mature human IGF-I; and the district court implicitly, but erroneously, upheld the Board's determination that Genentech had provided insufficient proof of the practical utility of its modified IGF-I because Genentech had failed to provide in vivo test results. Because we agree with Genentech that the district court did not give the count its broadest, reasonable interpretation and that Genentech's DNA construct is encompassed by the count, we need not reach the second issue concerning Chiron's enablement of the construct of the count. We also will not reach the utility issue because it was not decided by the district court and is therefore not before us. The sole issue before us is whether the proper construction of the count includes fusion proteins, such as modified IGF-I, or mature human IGF-I.

As Genentech points out, the count calls for a DNA construct which encodes certain proteins, but does not encompass the proteins expressed by the construct. The district court viewed the protein produced by the construct and the DNA construct as equal. As the district court stated, the issue of the DNA construct itself and the protein it produces are "inextricably intertwined since the DNA construct codes for and is defined by the protein it produces." Chiron, 1995 W.L. 450846, at *5.

Genentech asserts that against the backdrop of this erroneous equation of the DNA construct and the protein it produces, the district court improperly narrowed the count in the following manner.

First, the district court interpreted the DNA sequence coding for human IGF-I as defined by the count to mean that mature IGF-I, or the specific seventy amino acid protein, must be ultimately secreted from the transformed yeast cell containing the DNA construct of the count.

Second, the district court improperly limited the transitional phrase "comprising," which allows additional elements to be present as long as the named elements are present, to exclude additional DNA between the alpha-factor processing sequences and the human IGF-I sequence.

Third, Genentech asserts that the district court improperly gave a narrow meaning to "joined" by taking it out of the context of the phrase "joined in proper reading frame" and using a common dictionary definition of "joined" instead of one tailored to the biotechnical discipline.

By requiring a direct connection between the elements of the count "joined in proper reading frame," Genentech asserts that the clause "in proper reading frame" is rendered superfluous because, if the two DNA sequences coding for human IGF-I and the processing signal sequences are directly joined, they are necessarily in proper reading frame. Because the count must be interpreted so that every element of the count has meaning, Genentech asserts this interpretation is improper.

Genentech points out that the two examples provided by the district court to establish that a DNA construct can be directly joined, but not be in proper reading frame, are improper examples because such constructs would not meet the additional limitations of the count. For instance, the district court asserts that two DNA sequences can be directly connected, but not in proper reading frame because they overlap or one of the sequences is in reverse orientation. Genentech asserts that if the two sequences overlap, however, one of the DNA sequences will not be complete because the nucleotides in the section of DNA that overlaps will only be read once. Therefore, because the count calls for the DNA sequence for IGF-I and the alpha-factor processing sequences, such a construct would not meet all the elements of the count and the phrase "in proper reading frame" would be unnecessary to exclude such a construct from the count. Also, if one of the DNA sequences for one of the required elements is in reverse orientation, it would not encode that particular protein and would not meet the additional elements of the count. Again, the phrase "in proper reading frame" would be rendered superfluous.

Chiron asserts that the district court properly gave the count its plain meaning by properly interpreting the key phrases "human insulin-like growth factor-I," "coding for," "joined," and "comprising." As in the district court's interpretation of the count, Chiron agrees that the plain meaning of the word "coding" is a protein that results from the expression of the particular DNA that is encoded. Therefore, although the district court properly recognized that the count referred to the DNA construct and not the protein expressed, Chiron agrees with the district court that the two ideas are inextricably linked. To support this assertion, Chiron cites Genentech, Inc. v. Wellcome Foundation Ltd., 29 F.3d 1555, 31 USPQ2d 1161 (Fed. Cir. 1994) (phrase "human tissue plasminogen activator" includes only natural t-PA). Therefore, Chiron argues that, as the district court found, human IGF-I is a protein consisting of only seventy specific amino acids, and does not include the nine additional amino acids of the collagenase cleavage site; the phrase "joined in proper reading frame" necessitates that the elements of the DNA construct of the count must be joined so as to allow expression of the desired protein; and the term "comprising" is consistent with its interpretation of the word "joined" because DNA can be added before and after the joined elements of the count. Chiron asserts that even if the count is ambiguous and we must look to the Barr specification to determine the meaning of the count, the district court continues to have the correct interpretation of the count because the Barr application does not disclose a construct that would produce a protein other than naturally occurring human IGF-I.

D.

The pivotal issue on appeal is the proper construction of the count, which is a question of law. DeGeorge v. Bernier, 768 F.2d 1318, 1321, 226 USPQ 758, 760 (Fed. Cir. 1985). To construe the count we must look at the language as a whole and consider the grammatical structure and syntax. Credle v. Bond, 25 F.3d 1566, 1571, 30 USPQ2d 1911, 1915 (Fed. Cir. 1994).

In the absence of ambiguity, it is fundamental that the language of a count should be given the broadest reasonable interpretation it will support and should not be given a contrived, artificial, or narrow interpretation which fails to apply the language of the count in its most obvious sense. Only when counts are ambiguous may resort be had to the application where the counts originated, and this court does not look to the specification to determine whether there is an ambiguity.

In re Baxter, 656 F.2d 679, 686, 210 USPQ 795, 802 (CCPA 1981) (citations omitted).

We agree with both parties that no ambiguity exists in the count. The district court interpreted the count based on the protein encoded by the DNA construct. Therefore, the district court concluded that a DNA construct is outside of the scope of the count if, when incorporated into an expression plasmid and transformed into a yeast cell, it secretes a fusion protein instead of mature human IGF-I. The plain language of the count, however, militates against this conclusion. Although a close relationship exists between a DNA construct and the protein it encodes, the two are not equal. The count specifically defines a DNA construct, not the protein that is produced by expression from the construct. The specific elements enumerated in the count that necessarily must be included in any DNA construct within the count are a DNA sequence coding for the secretory leader, a processing signal sequence, and human IGF-I. No dispute exists that Genentech's DNA construct as described in the Lee application contains the complete DNA sequences for these three proteins in its DNA construct. The issue arises from the nine additional codons encoding the collagenase cleavage site inserted between the sequences coding for human IGF-I and the alpha-factor processing sequences. Therefore, the issue of claim construction is decided by the meaning of the phrase "joined in proper reading frame."

The term "reading frame" relates to the way in which a protein is expressed from a DNA construct. DNA is made up of a series of nucleotides. To express the DNA construct to yield a protein, the cell machinery reads nucleotides in sets of three, which are called triplets or codons, to incorporate specific amino acids into the protein. If an extra nucleotide is inserted or deleted from a DNA sequence, the series of triplets is changed and different amino acids are incorporated into the protein. The term "reading frame" describes this phenomenon. To maintain "proper reading frame," the triplets in the DNA sequence must be read so that the proper amino acids are incorporated into the resulting protein. See Karl Drlica, Understanding DNA and Gene Cloning: A Guide for the Curious, 33-36 (2d ed. 1992).

In the DNA construct of the count at issue here, "in proper reading frame" means that the nucleotides must be read in such a way that the seventy amino acids of human IGF-I are incorporated in the proper sequence in the expressed protein. Nothing exists in the definition of "in proper reading frame" to exclude nucleotides coding for additional amino acids at the beginning of the seventy amino acid IGF-I sequence. This interpretation of the count is consistent with the open-ended term "comprising." "Comprising" is a term of art used in claim language which means that the named elements are essential, but other elements may be added and still form a construct within the scope of the claim. In re Baxter, 656 F.2d at 686, 210 USPQ at 802.

The claim construction issue then becomes whether the term "joined" forecloses the possibility of additional nucleotides being inserted between the two joined elements, the alpha-factor processing sequences and human IGF-I sequence. We find that it does not.

The district court found that "joined" ordinarily means "connected." From this definition, the district court read into the term a requirement that the alpha-factor processing sequences and human IGF-I sequence should be directly joined with no intervening nucleotides. We hold, however, that this interpretation of the term "joined" is not the broadest, reasonable interpretation of the count. To be joined or connected does not necessitate a direct joining or connection. Nothing in the count restricts it to direct joining. Therefore, we hold that the count encompasses DNA constructs such as Genentech's, that have intervening nucleotide sequences between the alpha-factor processing sequences and the human IGF-I sequence, as long as the proper reading frame is maintained between the two joined sequences.

This is not inconsistent with our opinion in Wellcome Foundation, 29 F.3d 1555, 31 USPQ2d 1161. In that case we defined the phrase "human tissue plasminogen activator" in the context of the claims of the patent at issue in that case. Because we found the term ambiguous, we examined the specification for guidance as to the meaning and sought to avoid definitions upon which the PTO could not have relied when it issued the patent. We found through this method that the phrase "human tissue plasminogen activator" meant t-PA produced through recombinant DNA technology but having the same structure as natural t-PA. We found this definition "most consistent with the limited form in which the claims are drafted and the others were hopelessly overbroad." 29 F.3d at 1564, 31 USPQ2d at 1168.

This definition is not inconsistent with our holding here. Genentech's DNA construct contains codons for all seventy amino acids of the mature human IGF-I protein. Therefore, Genentech's construct contains the structure of natural IGF-I. Again, we are examining a DNA construct and not a protein. Therefore, the addition of codons at the beginning of the DNA sequence for human IGF-I is irrelevant.

III. Conclusion

We reverse the district court's determination that the sole count at issue does not encompass Genentech's DNA constructs. We remand the case for further proceedings in accordance with this opinion.

REVERSED and REMANDED.