Diurnal variations in angiostatin in human tear fluid: a possible role in prevention of corneal neovascularization.

PURPOSE: Although overnight eye closure is known to result in hypoxia and release of potent angiogenic factors, even prolonged eye closure does not result in corneal neovascularization. This suggests that the closed eye tear film may contain factors that can impede neovascularization. Closed eye tear fluid is known to contain proteases capable of converting plasminogen/plasmin to angiostatin and other angiostatin-like A-chain fragments which are potent inhibitors of angiogenesis. This study was designed to characterize open and closed eye tear fluid for the presence of these entities. METHODS: Open and closed eye tears were collected by microcapillaries from normal individuals. Tears were centrifuged and the supernatants analyzed by SDS-PAGE and western blotting. Membranes were probed with antibodies specific for the A-chain of plasmin and plasminogen and with antibodies specific for conformational domains on the smaller N terminal kringles 1-->4 and kringles 1-->3 fragments which are known angiogenesis inhibitors. Supernatants were also analyzed after fractionation by HPLC and binding to lysine sepharose 4B. The isolated fragments were identified based on size, lysine-binding capabilities, antigenic properties and by comparison with standards. RESULTS: Open eye tear fluid from all normal individuals contained low levels of plasminogen, but no detectable antigens consistent with free A-chain or angiostatins. Tears collected after overnight eye closure contained significant amounts of plasminogen, A-chain antigen and various A-chain fragments including kringles 1-->4 and kringles 1-->3 and most likely free kringle 5, all known to have anti-angiogenesis properties. These were often present in concentrations likely to be physiologically significant. In samples collected from an atopic subject, the concentration of angiostatins in CTF increased markedly during active phases of the disease reaching levels of several ng/microl. In these instances and in similar samples obtained from other atopic individuals experiencing active reactions, angiostatin was often detectable in basal-type tear fluid. CONCLUSION: A-chain fragments, which can inhibit angiogenesis, are often present at physiologically significant levels in human tear fluid collected after overnight eye closure. These fragments may play a role in preventing neovascularization in the hypoxic closed eye environment and may well be up regulated during inflammatory reactions.

Identification, origins and the diurnal role of the principal serine protease inhibitors in human tear fluid.

PURPOSE: Previous work identified polymorphonuclear leukocyte (PMN) elastase as the major caseinolytic entity in tears collected after overnight eye closure. This study was designed to identify the principal serine protease inhibitors (serpins) in tears and to determine their function in the regulation of PMN cell proteases on eye closure. METHODS: Reflex and closed eye tear samples were collected by microcapillary tube and centrifuged. After reflex and closed eye supernatants (R and C) were fractionated by HPLC, samples were subjected to casein zymography and reverse zymography. Western blots were utilized to screen tears and HPLC fractions for elastase, cathepsin G and proteinase-3 and to obtain semi-quantitative data on alpha 1-protease inhibitor (alp1), alpha 1-antichymotrypsin (alpha 1-Achy), secretory leukocyte protease inhibitor (SLPI), elafin and alpha 2-macroglobulin (alpha 2-M) as well as associated complexes and products. To confirm specificity of reactivity, samples were immunoprecipitated for a given protease or serpin and screened for the coprecipitation of interacting species. RESULTS: Although R fluid contains no caseinolytic activity, it contains low levels of serpin-like activity principally in the form of SLPI (5-10 ng/microliter). Lesser amounts of alpha 2-M, alpha 1-Achy and alp1 (approximately < 1-3 ng/microliter) are also evident. C fluid is associated with very high levels of PMN cell proteases along with a approximately 5-20-fold increase in the concentrations of all of the above inhibitors. Trace levels of elafin were also detected. The concentrations of rapid reacting inhibitors exceeded that of proteases, with SLPI, alpha 1-Achy and alp1 being the principal functional entities. In atypical samples, complexes of elastase and alpha 2-M were also encountered. CONCLUSIONS: SLPI, a known antimicrobial agent and an elastase and cathepsin G inhibitor, is the principal serpin in R fluid. C fluid is associated with a marked increase in the concentrations of an array of rapid reacting serpins capable of inhibiting all known PMN cell serine proteases. In the normal closed eye, the concentration of rapid reacting inhibitors always exceeds that of proteases with C fluid also containing a functional reserve of the slow reacting inhibitor alpha 2-M.

Polymorphonuclear leukocyte cells and elastase in tears.

PURPOSE: To characterize the effects that mode of sampling and overnight eye closure have on the nature of caseinolytic activity recovered in tear fluid. METHODS: Reflex, open and closed (R, O and C) eye tear fluids were collected by microcapillary tubes or from the inferior formix by Schirmer strip. Microcapillary collected samples were centrifuged and recovered cells cytochemically characterized and probed by immunofluorescence microscopy, or alternatively extracted in acidic PBS. Tear supernatants, pellets and Schirmer strip extracts were subjected to casein zymography or SDS-PAGE and immunoprobed for plasmin/plasminogen. To identify caseinolytic activity, samples were immunoprecipitated with antibodies to plasmin/plasminogen or to elastase, and the immunoprecipitated materials were subjected to zymographic analysis. RESULTS: Immunoblot assays revealed R and O samples contained low levels of plasminogen (approximately 1.1 micrograms/ml) and only trace levels of plasmin (< 0.1 ng/ml). Insufficient levels of caseinolytic activity were present to allow zymographic detection. Cytochemical analysis revealed that R and O pellets consisted almost exclusively of desquamated epithelium. Immunoblot analysis revealed that C fluid was associated with an increase in plasminogen and its partial conversion to plasmin (approximately 3.2 ng/ microliter), high molecular weight covalent complexes and degradative products. Zymographic analysis disclosed much greater caseinolytic activity than could be attributed to plasmin or its cleavage products. This consisted primarily of three bands (30-26 kDa) which were identified as polymorphonuclear leukocyte (PMN) cell elastase based on size and antigenicity. This is derived from PMNs recovered from the C pellet. Elastase could also be recovered from Schirmer strips from 90% of donors, provided that the strips were extracted in sample loading buffer. The activity was restricted to the portion of the strip that had been in contact with the ocular tissue. CONCLUSIONS: The main source of caseinolytic activity in C fluid is elastase. This arises from PMNs that undergo recruitment, activation and degranulation in the C environment. In contrast, the elastase recovered in Schirmer strip extracts is derived from intact PMNs that adhere to the strip during sample collection. This would suggest that PMN cells undergo a low level of recruitment into the open eye environment.

Synthesis of the templates for influenza virion RNA replication in vitro.

To elucidate the mechanism(s) of influenza viral RNA replication, we have developed an in vitro system in which the templates for viral RNA replication as well as the viral messenger RNAs (mRNAs) are synthesized. Because the synthesis of both the viral mRNAs and the template RNAs occurs in the nucleus of infected cells, we determined whether infected cell nuclei are active in the synthesis of these two types of transcripts in vitro. Nuclei isolated as early as 1-2 hr after infection catalyze the in vitro synthesis of both the viral mRNAs and template RNAs. The time course of appearance of these activities indicates that they most likely represent the transcriptional complexes functioning in vivo. Template RNA synthesis catalyzed by the nuclei in vitro is independent of concomitant protein synthesis; rather, it utilizes preformed proteins present in the nuclear preparations. This protein pool can be depleted by treating the infected cells with a protein synthesis inhibitor prior to the isolation of the nuclei, thereby rendering the nuclei inactive in template RNA synthesis in vitro. This activity can be restored by the addition of infected cell cytoplasmic extracts or of the high-speed supernatant fraction from these extracts. These results indicate that the cytoplasmic fraction from infected cells enables the viral transcription complex to continue transcription past the site at which termination occurs during viral mRNA synthesis and also suggest that this fraction enables the transcription complex to initiate transcription without the capped primer used in viral mRNA synthesis.

Selected host cell capped RNA fragments prime influenza viral RNA transcription in vivo.

Influenza viral RNA transcription in vitro is primed by capped RNA fragments cleaved from capped RNAs by a viral endonuclease. The present study was undertaken to determine whether the specificities of the viral endonuclease and transcriptase observed in in vitro studies are also observed in the infected cell. The NS (nonstructural) gene of influenza WSN virus was cloned in pBR322 by using a double-stranded DNA containing a cDNA copy of both virion RNA (vRNA) and in vivo viral mRNA. We determined the 5' terminal sequence of the particular NS viral mRNA molecule which was cloned and also the 5' terminal sequences of the entire population of in vivo NS viral mRNAs synthesized in two different cell lines. For the latter determination we used a restriction fragment from the cloned DNA for the reverse transcriptase-catalyzed extension of total in vivo viral mRNA. The results indicate that in vivo and in vitro viral RNA transcription are similar in two important respects: (i) transcription initiates not with an A residue directed by the 3' terminal U of the vRNA, but with a G residue directed by the 3' penultimate C of the vRNA; and (ii) capped RNA fragments containing a 3' terminal A residue are preferentially used as primers, therapy generating an AG sequence in the viral mRNA complementary to the 3' terminal UC of the vRNA. Actually, for in vivo transcription, a subset of A-terminated capped fragments, namely those containing a 3' penultimate C residue, are the preferred primers. The latter specificity had not been observed in previous in vitro studies.

Transcription antitermination during influenza viral template RNA synthesis requires the nucleocapsid protein and the absence of a 5' capped end.

The first step in the replication of influenza virion RNAs is the synthesis of full-length transcripts of these RNAs. The synthesis of these transcripts, or template RNAs, requires: unprimed initiation rather than the capped RNA-primed initiation used during viral mRNA synthesis, and antitermination at the polyadenylylation site used during mRNA synthesis. To determine the mechanism of template RNA synthesis, we prepared nuclear extracts from infected cells that were active in the synthesis of both template RNAs and viral mRNAs. By providing the dinucleotide ApG as primer, we circumvented the inefficient unprimed initiation catalyzed by these extracts and, as a consequence, were able to focus on the antitermination step. Antitermination, and hence template RNA synthesis, occurred when ApG but not a capped RNA was used as primer, indicating that the presence of a 5' capped end blocked antitermination at the 3' end of the transcript. Ultracentrifugation of the nuclear extract yielded a pellet fraction that contained viral nucleocapsids active in viral mRNA synthesis but not template RNA synthesis and a supernatant fraction that contained the antitermination factor. When the supernatant, which had essentially no activity by itself, was added to the pellet in the presence of ApG, template RNA synthesis was restored. Depletion experiments in which this supernatant was incubated with protein A-Sepharose containing antibodies to individual viral proteins demonstrated that the viral nucleocapsid protein was required for antitermination. The implications of these results for the control of viral RNA replication are discussed.

Location of the single gene for the large subunit of ribulosebisphosphate carboxylase on the maize chloroplast chromosome.

The structural gene for the large subunit of ribulosebisphosphate carboxylase (E.C. 4.1.1.39) in Zea mays is shown to be contained within a 2500 base pair sequence of chloroplast DNA. One copy of this DNA sequence is present in each circular maize chloroplast DNA molecule. It maps approximately 30,000 base pairs from the 5' end of the closest of two sets of rRNA genes and approximately 71,000 base pairs from the other set of rRNA genes.