Genentech, Inc.,plaintiff-Appellee v. Chiron Corporation,defendant-Appellant

220 F.3d 1345, 55 U.S.P.Q. 2d (BNA) 1636, 2000 U.S. App. LEXIS 18746, 2000 WL 1091450

• Court: Court of Appeals for the Federal Circuit
• Decided: August 4, 2000
• Docket: 99-1506
• Status: Published

Opinion

SCHALL, Circuit Judge.

This appeal stems from an interference between U.S. Patent Application Serial No. 06/506,078 (the “Lee application”), assigned to Genentech, Inc. (“Genentech”), and U.S. Patent Application Serial No. 06/922,199 (the “Barr application”), assigned to Chiron Corporation (“Chiron”). After the Board of Patent Appeals and Interferences (“Board”) awarded priority of invention to Chiron, see Lee v. Barr, Pat. Int. No. 102,208 (July 19, 1994) (Paper No. 176) (“Genentech I”), Genentech brought an action under 35 U.S.C. § 146 in the United States District Court for the Northern District of California. The district court conducted a bench trial, and, in due course, entered judgment in favor of Genentech. 1 See Genentech, Inc. v. Chi ron Corp., No. CV 94-03334 (N.D. Cal. June 24, 1999) (Amended Judgment). Chi-ron appeals this judgment, challenging the factual findings and legal conclusions of the district court. We reverse. We do so because we hold that Genentech failed to establish that the inventors on the Lee application reduced their invention to practice prior to the earlier effective filing date of the Barr application. Our holding is based on the conclusion that a non-inventor’s recognition of the utility of the invention of the Lee application did not inure to the benefit of the inventors.

BACKGROUND

I.

The subject matter of the interference relates to the recombinant production of human insulin-like growth factor-I (“IGF-I”). IGF-I is a growth-promoting protein. IGF-I works by binding to IGF-I receptors, which are proteins embedded in cell membranes. Binding between IGF-I and IGF-I receptors causes a signal to be transmitted across the cell membrane. This signal activates the intracellular portion of the receptor, which triggers growth-related processes within the cell. The signal is believed to result from a change in the conformation (the three-dimensional shape) of the receptor that occurs when IGF-I binds to the receptor.

Both Genentech and Chiron wanted to use the organism Saccharomyces, commonly known as baker’s yeast, to make IGF-I recombinantly. Yeast makes and secretes a protein known as alpha-factor during its reproductive cycle. When yeast makes alpha-factor, it first makes a precursor protein that comprises a secretory leader sequence and four .repeating units of processing signal sequences and alpha-factor sequences:

leader — processing signal — alpha-factor — processing signal — alpha-factor— processing signal — alpha-factor—processing signal — alpha-factor

The leader sequence is a signal for secretion, i.e., the yeast cell recognizes this sequence and secretes the precursor protein from the cell: Each processing signal sequence in the precursor protein is a recognition site for a proteolytic enzyme— an enzyme that acts at the sequence to cleave (cut) the protein. Thus, when the yeast makes a precursor protein, the precursor is secreted from the cell and cleaved by an enzyme to yield four copies of the mature alpha-factor protein. This is called an expression-secretion system because the yeast expresses the alpha-factor DNA — it makes the precursor protein— and then secretes the protein from the cell. Expression-secretion systems are an advantageous way to make recombinant proteins because the proteins can be recovered without destroying the cells that produce them.

Genentech and Chiron sought to adapt the yeast alpha-factor DNA to make IGF-I. Specifically, they used a DNA construct comprising DNA encoding the alpha-factor secretory leader sequence and the alpha-factor processing signal sequence and DNA encoding IGF-I:

leader — processing signal — IGF-I

When yeast is transformed with this DNA, the yeast first makes a precursor protein comprising the leader sequence, the processing signal sequence and the IGF-I sequence. This protein is then processed by enzymes to cleave the leader and spacer sequences, yielding the mature IGF-I protein.

The interference count defines the invention at issue as follows:

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 sequences.

Genentech I, slip op. at 2. In Genentech II, we construed this count as requiring the following elements: (1) a DNA sequence coding for the yeast alpha-factor secretory leader sequence; (2) a processing signal sequence; and (3) a DNA sequence coding for IGF-I. See Genentech II, 112 F.3d at 501, 42 USPQ2d at 1613.

The construct that is alleged to support Genentech’s reduction to practice of the count has an additional DNA sequence between the DNA encoding the processing signal sequence and the. DNA encoding IGF-I. See id. at 498, 42 USPQ2d at 1610. That sequence codes for a nine amino acid long collagenase cleavage site. See id. That is, it codes for an amino acid sequence that is cleaved by the enzyme col-lagenase. Thus, the Genentech construct has the following structure:

leader — processing signal — collagenase cleavage site — IGF-I

In Genentech II, we held that the count was broad enough to encompass this construct. See id. at 501, 42 USPQ2d at 1614.

When the Genentech construct is made and processed by the yeast cells, the leader and processing signal sequences are cleaved by the yeast enzymes, but the collagenase cleavage site sequence is not cleaved from the IGF-I sequence. Thus, the final product obtained from yeast transformed with the Genentech DNA construct is a fusion protein consisting of the collagenase cleavage site and the IGF-I sequence. Genentech intended to subject this fusion protein to further processing in order to obtain only the IGF-I sequence (the mature IGF-I protein), but its work in this regard was not successful.

II.

The Board determined the priority of invention to the count in Interference No. 102,208. Because the Barr application had an earlier effective filing date than the Lee application, Barr (Chiron) was the senior party in the interference. See 37 C.F.R. § 1.657(a). Lee (Genentech) therefore had the burden of establishing by a preponderance of the evidence that he had reduced the invention to practice prior to the effective filing date of the Barr application (the “critical date”). 2 See id. § 1.657(b).

The Board determined that Lee had made a construct within the scope of the count. See Genentech I, slip op. at 8. As discussed above, this construct comprised DNA encoding the alpha-factor secretory leader sequence, DNA encoding the alpha-factor processing signal sequence, DNA encoding a collagenase cleavage site, and DNA encoding IGF-I. Because this construct did not result in a known compound, IGF-I, but in a novel fusion protein consisting of the collagenase cleavage site and IGF-I, the Board concluded that there could be no reduction to practice of the invention until a practical utility for the fusion protein had been established. See id. at 16. The Board found that the only contemplated utility for the fusion protein was use as a growth-promoting therapeutic agent, see id. at 17; accordingly, the Board considered whether Lee had established, prior to the critical date, that the fusion protein had growth-promoting activity. Although the Board considered several tests performed on the fusion protein, see id. at 17-18, only one, the radioreceptor assay (“RRA test”), is relevant to this appeal.

The RRA test is an in vitro assay that determines the ability of a protein to competitively bind to a receptor. The RRA test at issue here evaluated the ability of the fusion protein to bind to IGF-I receptors in a placental membrane. The test used both the fusion protein and radiola-beled mature IGF-I. The radiolabeled mature IGF-I was contacted with the placental membrane, and, after a period of time, the unbound material was washed away and the radioactivity level of the membrane-bound material was measured. Then, the fusion protein was introduced to the system and, after a period of time, unbound material was washed away and the radioactivity level of the membrane-bound material was again measured. The second radioactivity level was found to be lower than the first, indicating that the fusion protein had displaced some of the radiolabeled mature IGF-I and, therefore, that the fusion protein had bound to the IGF-I receptors specifically and competitively. The fusion protein therefore yielded “positive” results in the RRA test.

The Board determined that the positive RRA results were “not sufficient to establish biological activity with any reasonable certainty .... [because] the testimony of Dr. Ullrich [an inventor on the Lee application] makes clear that there was no prior art recognized correlation between the RRA assay and a specific in vivo therapeutic use.” Id. at 18. The Board also cited the testimony of Dr. Rutter and Dr. Spencer, Chiron’s expert witnesses, for the proposition that a fusion protein could have a three-dimensional structure that binds to an IGF-I receptor, as indicated by positive RRA test results, but the structure might be such that the binding does not result in the desired biological response. That is, a protein could bind to an IGF-I- receptor, but the resulting conformational change in the receptor might not trigger the growth-promoting signal that is triggered when- IGF-I binds the receptor. Such a protein would be an IGF-I antagonist because it would interfere with the activity of IGF-I by blocking IGF-I receptors without promoting growth itself.

The Board therefore held that Lee had not established a reduction to practice of the invention prior to the critical date, and awarded priority of invention to Barr. See id. at 19, 22. .

III.'

Genentech challenged the Board’s decision in an action under 35 U.S.C. § 146 3 in the United States District Court for the Northern District of California. In that action, both Genentech and Chiron presented new evidence on the correlation between positive RRA test results and biological activity. This new evidence included live testimony from Dr. Ullrich, Dr. Hintz (a paid consultant for Genentech), and séveral other experts.

The court determined that Genentech had to establish three things in order to demonstrate its prior reduction to practice of the invention: (1) that its inventors had made an embodiment of the count prior to the critical date; (2) that sufficient- tests had been conducted prior to the critical date to demonstrate that the invention worked for its intended purpose; and (3) that the inventors recognized, prior to the critical date, that the invention worked for its intended purpose. See Genentech, Inc. v. Chiron Corp., No. CV 94-03334, slip op. at 33 (N.D.Cal. Feb. 4, 1999) (Findings of Fact and Conclusions of Law) (“Genentech III”).

The court found that Genentech had established that its inventors had made a DNA construct within the scope of the count. The court recognized that the Board had made such a finding, and noted that Chiron had failed to present any evidence demonstrating that the Board’s finding was erroneous. See id. at 34.

The court determined that the intended purpose of the invention, the DNA construct, was to transform yeast cells so that the yeast cells would “secrete a form of biologically active human IGF-I, either as the mature IGF-I protein or in a modified form such as a fusion protein.” See id. at 35. The court therefore stated that to demonstrate that the invention worked for its intended purpose, Genentech had to establish that.“the invention produced an IGF-I-like fusion protein” and that “the fusion protein had a practical utility, i.e., that it exhibited IGF-I biological activity.” Id.

The court found that Genentech had established that the invention produced an IGF-I-like fusion protein. See id. at 37. Specifically, the court concluded that the DNA sequence of the DNA construct, the approximate molecular weight of the fusion protein, and test results demonstrating that the fusion protein bound to IGF-I antibodies and IGF-I receptors were sufficient to show that Genentech’s construct produced an IGF-I-like fusion protein. See id.

The court determined, contrary to the Board’s holding, that the practical utility of the invention had been established. The court considered the evidence of a correlation between positive RRA test results and in vivo biological activity, and concluded that, by the critical date, “[s]pecific binding in an RRA was known ... by those skilled in the art to be reasonably correlated with the in vivo biological activity of IGF-I.” Id. at 40. The court cited the testimony of Dr. Hintz, Dr. Ravtech (a paid consultant for Genentech), and Dr. Ullrich as supporting its conclusion. See id. The court recognized the Board’s concern with the RRA test — that the test did not eliminate the possibility that the fusion protein was an IGF-I antagonist — but determined that, under Fujikawa v. Wattanasin, 93 F.3d 1559, 39 USPQ2d 1895 (Fed.Cir.1996), “test results need not absolutely prove that the compound is pharma-cologically active.” Genentech III, slip op. at 39 (citing Fujikawa, 93 F.3d at 1564, 39 USPQ2d at 1899). The court concluded that the tests Genentech had conducted on the fusion protein, including the RRA test, were sufficient to establish the practical utility of the fusion protein. See id.

The court initially ruled against Genen-tech on the third prong of its reduction to practice test, finding that Genentech had failed to prove that its inventors had recognized, prior to the critical date, that the fusion protein worked for its intended purpose. See id. at 42. The court found that two of the inventors, Dr. Ullrich and Mr. Lee, knew prior to the critical date that the fusion protein had yielded “positive” results in the RRA test. See id. at 43. The court determined, however, that there was no evidence that the inventors understood the positive RRA results to indicate that the fusion protein had pharmacological activity. See id. at 43-45. Rather, the court concluded, the evidence showed only that the inventors interpreted the results as indicating that there was an IGF-I-like substance in the material that was tested. See id.

The court reconsidered its judgment, however, and ultimately entered an amended judgment in favor of Genentech. The court determined that, although the inventors had not recognized the practical utility of the invention prior to the critical date, Dr. Hintz, the person who had conducted the RRA test, had. See Genentech, Inc. v. Chiron Corp., No. CV 94-03334 (N.D. Cal. June 22, 1999) (Order Granting PL’s Mot. to Amend the Court’s Conclusions of Law and J.) (“Genentech IV”). Because it was “undisputed” that Dr. Hintz “was engaged as an agent of Genen-tech for purposes of conducting tests on the material provided,” id. at 8, the court concluded that “Dr. Hintz’s .appreciation of the significance of the results of his tests ... inures to the benefit of the Genentech inventors for purposes of establishing a reduction to practice.” Id. The court therefore held that Genentech had reduced the invention to practice in October of 1982. See id. at 9.

The court considered Chiron’s evidence of conception, and determined that Chi-ron’s date of conception was later than Genentech’s reduction to practice. Accordingly, the court awarded priority of invention to Genentech. See id. at 22. Chiron appeals the court’s judgment in favor of Genentech. We have jurisdiction pursuant to 28 U.S.C. § 1295(a)(4)(C).

DISCUSSION

Chiron challenges the district court’s judgment in favor of Genentech on several grounds. It argues first that the court should háve given more deference to the Board’s finding that the positive RRA test results did not establish the practical utility of the fusion protein, and should not have substituted its judgment for that of the Board. Chiron also asserts that several of the court’s factual findings were clearly erroneous, including the finding that Genentech had made a fusion protein within the scope of the count, the finding that the Genentech inventors intended to use the fusion protein as a pharmacological agent, the finding that the Genentech inventors had established, through the positive RRA test results, that the fusion protein has growth-promoting activity, and the finding that Dr. Hintz had contemporaneously recognized that the fusion protein has growth-promoting activity. Chiron further contends that the district court committed legal error when it relied on the inventors’ uncorroborated evidence to support its conclusion that the sample that yielded the positive RRA test results fell within the scope of the count. Finally, Chiron asserts that, on the facts of this case, Dr. Hintz’s alleged recognition of the growth-promoting activity of the fusion protein should not inure to the benefit of Genentech.

We note at the outset that the district court was entitled to decide de novo whether Genentech had.established a practical utility for the fusion protein. The parties’ briefs were submitted before we decided Winner International Royalty Corp. v. Wang, 202 F.3d 1340, 53 USPQ2d 1580 (Fed.Cir.2000). In Winner, we held “that the admission of live testimony on all matters before the Board in a section 146 action ... makes a factfinder of the district court and requires a de nova trial.” Id. at 1347, 53 USPQ2d at 1585. Although it does not appear that the district court here heard live testimony on all issues decided by the Board, the court did hear live testimony on the issue of whether Genentech had established a practical utility for the fusion protein. Specifically, the court heard live testimony on this issue from Dr. Ullrich, Dr. Hintz, Dr. Ravtech, Dr. Spencer, Dr. Rutter, and two other experts. Under Winner, live testimony on the issue of practical utility makes the district court a factfinder on that issue, and requires the court to decide that issue de novo. Accordingly, the district court was not required to give deference to the Board’s finding that Genentech had not established a practical utility for the fusion protein.

Although we have carefully considered the other arguments raised by Chi-ron; because we find the inurement issue to be dispositive, we focus our attention on that issue. “We review the district court’s factual findings for clear error and its conclusions of law de novo.” Winner, 202 F.3d at 1344-45, 53 USPQ2d at 1583. We therefore will consider whether the district court’s factual finding that Dr. Hintz contemporaneously recognized the growth-promoting activity of the fusion protein was clearly erroneous and whether the court’s legal conclusion that Dr. Hintz’s recognition of the utility of the fusion protein inures to the benefit of Genentech was correct.

III

Chiron argues that the record does not support the district court’s finding that Dr. Hintz recognized, prior to the critical date, that the fusion protein has growth-promoting activity. Specifically, Chiron emphasizes that Dr. Hintz never stated that he had recognized the growth-promoting activity of the fusion protein in 1982. Instead, Chiron asserts, he only testified that he would have interpreted positive RRA test results as indicating that the fusion protein would be expected to have pharmacological activity. Chiron contends that this testimony is merely his current opinion about what -the test results would have indicated at' the time, and is not evidence of his contemporaneous appreciation that the fusion protein has' growth-promoting activity.

In response, Genentech argues that Dr. Hintz’s testimony indeed establishes his contemporaneous recognition of the growth-promoting activity of the fusion protein. Genentech emphasizes that Dr. Hintz had been directed to testify in the context of 1982, without reference to later developments. Genentech also cites Dr. Hintz’s testimony that he did not repeat the RRA test because he viewed the positive results as demonstrating that the fusion protein has activity. It argues that this testimony supports the court’s finding that Dr. Hintz had appreciated the growth-promoting activity of the fusion protein at that time.

The district court heard the following testimony from Dr. Hintz:

Q. Do the results of this test [the RRA test] tell anything about whether the Genentech material would be expected to have pharmacologic activity?

A. In my opinion, at that time and still, the test results indicate that it would be expected to have pharmacologic activity on the basis of its ability to bind the receptor site. For IGF specifically.

The Court. Can you explain what you mean by “pharmacologic activity”?

The Witness. That if you were to make enough of this and give it to an animal or a human ... you could make them grow.

Q. You testified that you only did the RRA once. Is there a reason why you didn’t repeat that assay?

A. I thought that it was clear that there was activity in that Lee ... construct ....

Trial Tr. at 390, 437. Dr. Hintz also testified that, in 1982, his opinion was that positive RRA test results reflected biological activity. See, e.g., id. at 410-11, 439. On this record, we cannot say that the district court’s finding that Dr. Hintz recognized the pharmacological activity of the fusion protein prior to the critical date is clearly erroneous.

Chiron argues that, even if Dr. Hintz did recognize the practical utility of the fusion protein, his recognition should not inure to the benefit of Genentech. Chiron contends that the Genentech inventors did not intend or expect the fusion protein to be tested for growth-promoting activity. It asserts that the inventors’ only interest in the fusion protein was as a precursor for the mature IGF-I protein, and that the inventors were not interested in the biological activity of the fusion protein. Accordingly, Chiron argues, the inventors only wanted Dr. Hintz to determine whether the fusion protein had been expressed and secreted by yeast, and not whether it had a practical utility.

Genentech responds that because Dr. Hintz was engaged as Genentech’s agent to test the fusion protein, his recognition of the practical utility of the fusion protein inures to Genentech’s benefit. Genentech argues that as long as the experiments that demonstrated utility were performed at the direction of the inventors, the results of those experiments inure to the benefit of the inventors. Genentech contends that the inventors submitted the samples to Dr. Hintz to test for the intended utility of the invention; therefore, Dr. Hintz’s recognition that the fusion protein would work for that purpose inures to the inventors’ benefit, and completes the reduction to practice of the invention. Alternatively, Genentech argues that the reduction to practice was complete when Dr. Hintz conveyed the positive RRA test results to Dr. Ullrich.

“In order to establish an actual reduction to practice, the inventor must prove that: (1) he constructed an embodiment ... that met all the limitations of the interference count; and (2) he determined that the invention would work for its intended purpose.” Cooper v. Goldfarb, 154 F.3d 1321, 1327, 47 USPQ2d 1896, 1901 (Fed.Cir.1998). “When testing is necessary to establish utility, there must be recognition and appreciation that the tests were successful for reduction to practice to occur.” Estee Lauder Inc. v. L’Oreal, S.A., 129 F.3d 588, 594-95, 44 USPQ2d 1610, 1615 (Fed.Cir.1997). The district court determined that the Genentech inventors did not understand the positive RRA test results to demonstrate the practical utility of the fusion protein. See Genentech III, slip op. at 47. The court concluded that the inventors understood the positive test results to indicate that there was “a significant amount of IGF-I-like fusion protein” in the samples, Genentech III, slip op. at 42, but did not understand the results to indicate the “pharmacological activity of the fusion protein,” id. at 45.

Genentech argues that Dr. Ullrich did understand from the RRA test results that the fusion protein worked for the intended purpose of the invention. In support of this argument, it cites deposition testimony of Dr. Ullrich that he knew from the positive RRA test results that the fusion protein “was very likely to be biologically active.” However, Dr. Ullrich testified before the district court that he did not know in 1982 that there was a correlation between the RRA test and a specific in vivo therapeutic use, and that the RRA test was just one test “that together with other results would ... establish the biological activity of the molecule.” Trial Tr. at 191. When we consider the record as a whole, we do not find the district court’s conclusion that Dr. Ullrich had not appreciated the practical utility of the fusion protein to be clearly erroneous. We therefore must reject Genentech’s argument that its reduction to practice was complete when Dr. Hintz reported the positive RRA test results to the inventors. Accordingly, Gen-entech can prevail only if Dr. Hintz’s recognition of the growth-promoting activity of the fusion protein inures to its benefit.

“Inurement involves a claim by an inventor that, as a matter of law, the acts of another person should accrue to the benefit of the inventor.” Cooper, 154 F.3d at 1831, 47 USPQ2d at 1904. In Estee Lauder, we suggested that the requirement that an inventor mu'st know that the invention is useful might be satisfied when an agent of the inventor obtains such knowledge. See Estee Lauder, 129 F.3d at 593, 44 USPQ2d at 1614. However, our case law does not clearly define the circumstances under which this type of inurement may occur. We did not directly address this issue in Estee Lauder because we determined that no one had recognized the utility of Estee Lauder’s invention prior to the critical date. See id. at 595, 44 USPQ2d at 1615. In Reese v. Hurst, 661 F.2d 1222, 211 USPQ 936, 941, 943 (CCPA 1981), our predecessor court affirmed a Board decision that a reduction to practice had occurred on the date that test results were obtained instead of on the later date when the results were conveyed to the inventor; however,’the court did not explain this aspect of its decision. See Estee Lauder, 129 F.3d at 594, 44 USPQ2d at 1615 (noting that whether the invention at issue in Reese was reduced to practice on the testing date or on the date when the results were conveyed to the inventor “was entirely immaterial to the decision” because both dates were prior to the critical date, and stating that, to the extent that Reese can be read as “imposing a rule that the. successful testing date is the date of reduction to practice, even when success is not known until. later[,] .... it stands alone among our precedent”). By the same token, the court in Gunter v. Stream, 573 F.2d 77, 78, 197 USPQ 482, 483 (CCPA 1978), stated without explanation that tests conducted by fellow employees of the inventor were “proof of an actual reduction tó practice which is attributable to [the inventor],” even though the inventor “took no part in th[at] phase” of development.

Two decisions of the Board illustrate when an inventor can benefit from a non-inventor’s recognition that the invention works for its intended purpose. In D’Silva v. Drabek, 214 USPQ 556, 1981 WL 46261 (BPAI 1981), the Board rejected the argument that the inventor had to establish that the relevant test results had been communicated to him. D’Silva had submitted a sample to the Biological Evaluation Group of his company for testing for pesticidal activity. See id. at 559. The Board reasoned that because “D’Silva had a conception of the invention ... coupled with his expectation of insecticidal activity, the work of the Biological Evaluation Group in testing the compound inured to D’Silva’s benefit whether he knew about it or not.” Id. at 563. Similarly, in De Solms v. Schoenwald, 15 USPQ2d 1507, 1509, 1990 WL 354513 (BPAI 1990), the Board stated that “it is unnecessary for the test results to be communicated to De Solms [the inventor] to establish a reduction to practice.” The inventor had submitted a compound “for evaluation of its activity in inhibiting the carbonic anhy-drase enzyme.” Id. at 1509. The Board explained that because “De Solms had a conception of the invention ... coupled with her expectation of carbonic anhydrase enzyme inhibiting activity, the work of' Sondey in evaluating the compound accrued to De Solms’ benefit whether she knew about it or not.” Id.

From these cases, we glean at least three requirements that must be met before a non-inventor’s recognition of the utility of an invention can inure to the benefit of the inventor. First, the inventor must have conceived of the invention. Second, the inventor must have had an expectation that the embodiment tested would work for the intended purpose of the invention. Third, the inventor must have submitted the embodiment for testing for the intended purpose of the invention. Applying these requirements to the case before us, we conclude that Dr. Hintz’s recognition of the growth-promoting activity of the fusion protein does not inure to Genenteeh’s benefit.

There is no evidence of record that the inventors submitted the fusion protein for testing for the intended purpose of the invention, which is promoting growth. Rather, as the district court found, when the inventors asked Dr. Hintz to test the fusion protein samples for IGF-I “activity,” they were asking him to test the samples for “the presence of an IGF-I-like substance.” Id. at 45. They were not, as Genentech asserts, asking him to test the sample for the intended purpose of the invention. Because the inventors did not submit the fusion protein samples to Dr. Hintz for testing for growth-promoting activity, his uncommunicated recognition that the fusion protein has that activity does not inure to their benefit. The district court therefore erred when it concluded that the Genentech inventors are entitled to the benefit of Dr. Hintz’s discovery that the fusion protein worked “in a manner ... [the inventors] had not anticipated.” Genentech IV, slip op. at 9 (emphasis added).

CONCLUSION

For the foregoing reasons, the judgment awarding priority of invention to Genentech is

REVERSED.

COSTS

Each party shall bear its own costs.

1 . The court initially granted summary judgment in favor of Chiron. See Genentech, Inc. v. Chiron Corp., No. CV 94-03334 (N.D.Cal. July 28, 1995) (Order). Genentech appealed to this court, and we reversed and remanded, based on our determination that the district court had construed the count too narrowly. See Genentech, Inc. v. Chiron Corp., 112 F.3d 495, 42 USPQ2d 1608 (Fed.Cir.1997) (“Gen entech II”). It is the court’s decision on remand that is at issue in this appeal.

2 . The Board precluded Barr from establishing an earlier date of invention because he failed to comply with the requirements of 37 C.F.R. § 1.622 regarding preliminary statements in an interference proceeding. See Genentech I, slip op. at 19.

3 . This statute provides in relevant part that "[a]ny party to an interference dissatisfied with the decision of the Board of Patent Appeals and Interferences on the interference, may have remedy by civil action....” 35 U.S.C. § 146 (1994).