`
`http://science.sciencemag.org/
`
`Downloaded from
`
`not related to an inability of deoxycholate
`to release MTP from the microsomes, we
`sonicated the microsomes from one abeta-
`lipoproteinemic subject for 5 min after de-
`tergent treatment. Sonication releases TG
`transfer activity with an efficiency compa-
`rable to that of detergent treatment. Even
`after sonication, no TG transfer activity
`was detected.
`To demonstrate that the lack of detect-
`able TG transfer activity in the abetali-
`poproteinemic individuals was not related
`to an inability to detect activity in cells that
`contain large intracellular fat droplets, as
`occur in abetalipoproteinemia, we mea-
`sured TG transfer activity in biopsy samples
`from a subject with Anderson's disease (also
`referred to as chylomicron retention dis-
`ease) (11) and a subject with homozygous
`hypobetalipoproteinemia (12). In these two
`genetic diseases, defects occur in the assem-
`bly or secretion of chylomicrons, and affect-
`ed individuals have large fat droplets in
`their enterocytes, analogous to individuals
`with abetalipoproteinemia.
`In addition,
`TG transfer activity was measured in an
`intestinal biopsy sample taken from a nor-
`mal subject who had not fasted before the
`biopsy. In all three individuals, TG transfer
`activity comparable to that of the control
`subjects was measured (Table 1), confirm-
`
`25
`
`20
`
`15
`
`10
`
`ag 5
`
`= O
`Ch
`S%_ 25
`
`o 20
`I-
`
`.B
`
`15 I
`
`10 h
`
`5
`
`80
`
`O I F
`0°
`
`20
`
`ga-g--
`60
`
`80
`
`40
`Protein (rp)
`Fig. 1. TG transfer activity in (A) five normal and
`(B) four abetalipoproteinemic individuals. TG
`transfer activity was measured in homogenized
`intestinal biopsy samples. Results are ex-
`pressed as the percentage of the total [14C]TG
`transferred from donor to acceptor membranes
`in a 1-hour assay as a function of the amount of
`intestinal protein that was treated with deoxy-
`cholate to release TG transfer activity.
`
`9119219~
`
`ig..9999RRI
`
`Absence of Microsomal Triglyceride Transfer
`Protein in Individuals with Abetalipoproteinemia
`
`John R. Wetterau,* Lawrence P. Aggerbeck,
`Marie-Elisabeth Bouma, Claude Eisenberg, Anne Munck,
`Michel Hermier, Jacques Schmitz, Gerard Gay, Daniel J. Rader,
`Richard E. Gregg
`Abetalipoproteinemia is a human genetic disease that is characterized by a defect in the
`assembly or secretion of plasma very low density lipoproteins and chylomicrons. The
`microsomal triglyceride transfer protein (MTP), which is located in the lumen of microsomes
`isolated from the liver and intestine, has been proposed to function in lipoprotein assembly.
`MTP activity and the 88-kilodalton component of MTP were present in intestinal biopsy
`samples from eight control individuals but were absent in four abetalipoproteinemic sub-
`jects. This finding suggests that a defect in MTP is the basis for abetalipoproteinemia and
`that MTP is indeed required for lipoprotein assembly.
`
`Abetalipoproteinemia is an autosomal-re-
`cessive disease that is characterized by a
`virtual absence of plasma lipoproteins that
`contain apolipoprotein B (apoB) and by
`low plasma concentrations of triglyceride
`(TG) and cholesterol (1). These abnormal-
`ities are the result of a genetic defect in the
`assembly or secretion of very low density
`lipoproteins (VLDLs) in the liver and of
`chylomicrons in the intestine, resulting in
`retinitis pigmentosa, spinocerebellar degen-
`eration with ataxia, and a bleeding diathe-
`sis secondary to malabsorption of fat-soluble
`vitamins. The molecular basis for the pri-
`mary defect in abetalipoproteinemia has
`not been determined. TG, phospholipid,
`and cholesterol synthesis are not impaired
`(1), and linkage between the apoB gene
`and abetalipoproteinemia has been exclud-
`ed by restriction fragment length polymor-
`phism (RFLP) analysis in several families
`(2, 3).
`We investigated the possibility that a
`defect in the microsomal TG transfer pro-
`tein (MTP) may be the proximal cause of
`abetalipoproteinemia. MTP is a soluble
`protein present in the lumen of microsomes
`J. R. Wetterau and R. E. Gregg, Department of Meta-
`bolic Diseases, Bristol-Myers Squibb, Princeton, NJ
`08543-4000.
`L. P. Aggerbeck, Centre de Genetique Mol6culaire,
`Centre National de Ia Recherche Scientifique, 91198
`Gif-sur-Yvette, France.
`M.-E. Bouma and C. Eisenberg, FacultM de M6decine
`Xavier Bichat, Unite 327 Institut National de Ia Sante et
`de Ia Recherche M6dicale, 75018 Paris, France.
`A. Munck, Service Gastroent6rologie Nutrition Podia-
`triques, H6pital Robert Debr6, 75019 Paris, France.
`M. Hermier, Clinique Modicale Infantile B, Hepatogas-
`troenterologie et Nutrition Infantiles, H6pital Edouard
`Herriot, 69437 Lyon, France.
`J. Schmitz, Departement de Podiatrie, H6pital des
`Enfants Malades, 75015 Paris, France.
`G. Gay, Service de M6decine Interne, H6pital Saint
`Nicolas, 55107 Verdun, France.
`D. J. Rader, Molecular Disease Branch, National
`Heart, Lung and Blood Institute, National Institutes of
`Health, Bethesda, MD 20892.
`*To whom correspondence should be addressed.
`
`isolated from liver and intestine (4). It
`mediates the transport of TG, cholesteryl
`ester, and phosphatidylcholine (PC) be-
`tween membranes (5). The ability of MTP
`to transport TG between membranes, to-
`gether with its tissue distribution and sub-
`cellular location, has led to the suggestion
`that MTP functions in the assembly of
`plasma lipoproteins (4).
`MTP has been purified from bovine liver
`and characterized (5). It is a heterodimer of
`58- and 88-kD peptides (6). Characteriza-
`tion of the 58-kD component indicated
`that it is the previously described multi-
`disulfide
`functional
`protein,
`protein
`isomerase (PDI) (7). The role of PDI in the
`transfer protein complex is not known. At a
`minimum, PDI appears to be necessary to
`maintain the structural integrity of the
`transfer protein (8), but a larger role cannot
`be excluded. Because PDI by itself does not
`have lipid transfer activity, the 88-kD sub-
`unit is either the active component or it
`confers transfer activity to the protein com-
`plex.
`MTP activities in duodenal or duodenal-
`jejunal junction biopsy samples obtained
`from abetalipoproteinemic (9) and normal
`control subjects after an overnight fast were
`compared. Intestinal biopsy tissue was ho-
`mogenized and treated with deoxycholate
`to release TG transfer activity from the
`microsomal fraction (10). The membrane
`fractions were removed by centrifugation,
`and TG transfer activity was measured in
`the supermatants. In biopsy samples from all
`five normal control subjects tested (Fig.
`1A), TG transfer activity (10) was readily
`detectable. TG transfer activity was not
`detected ('5% of the mean of the normal
`subjects) in the biopsy tissue from any of
`the four abetalipoproteinemic subjects (Fig.
`lB and Table 1). To demonstrate that the
`lack of detectable TG transfer activity in
`individuals with abetalipoproteinemia was
`
`SCIENCE * VOL. 258
`
`*
`
`6 NOVEMBER 1992
`
`999
`
`1 of 4
`
`PENN EX. 2056
`CFAD V. UPENN
`IPR2015-01835
`
`
`
`up
`
`:. a
`
`g
`
` on June 6, 2016
`
`http://science.sciencemag.org/
`
`Downloaded from
`
`that lipoproteins isolated from a rat liver
`Golgi fraction were consistently larger than
`those isolated from an ER fraction, which
`suggests the addition or transfer of lipid
`molecules to the nascent particles. The
`progressive addition of lipid to a developing
`lipoprotein particle was also demonstrated
`in the pulse-chase studies of Janero and
`Lane (18) and Bostrom et al. (19).
`The absence of the 88-kD component of
`MTP in individuals with abetalipoproteine-
`mia could be attributable either to its down-
`regulation to a nondetectable level or to a
`genetic defect in MTP or a factor that
`controls MTP concentration. Although
`MTP could be down-regulated in response
`to the cells not secreting lipoproteins, this
`explanation is unlikely because MTP con-
`centrations were normal in the subjects
`with Anderson's disease or homozygous hy-
`pobetalipoproteinemia, diseases in which
`enterocytes do not secrete lipoproteins.
`Given that all other known aspects of
`lipoprotein synthesis and assembly-the ex-
`pression of a normal apoB gene (2, 3, 20),
`as well as TG, phospholipid, and cholester-
`ol synthesis-are not impaired in abetali-
`poproteinemic subjects, it is likely that the
`proximal cause of abetalipoproteinemia is a
`genetic defect in the 88-kD component of
`MTP or in the regulation of its synthesis or
`degradation.
`
`REFERENCES AND NOTES
`1. J. P. Kane and R. J. Havel, in The Metabolic Basis
`of Inherited Disease, C. R. Scriver, A. L. Beaudet,
`W. S. Sly, D. Valle, Eds. (McGraw-Hill, New York,
`ed. 6, 1989), pp. 1139-1164.
`2. P. J. Talmud et al., J. Clin. Invest. 82, 1803 (1988).
`3. L. S. Huang et al., Am. J. Hum. Genet. 46, 1141
`(1 990).
`4. J. R. Wetterau and D. B. Zilversmit, Biochim.
`Biophys. Acta 875, 610 (1986).
`_ , Chem. Phys. Lipids 38, 205 (1985).
`5.
`6. J. R. Wetterau, L. P. Aggerbeck, P. M. Laplaud, L.
`R. Mclean, Biochemistry 30, 4406 (1991).
`7. J. R. Wetterau, K. A. Combs, S. N. Spinner, B. J.
`Joiner, J. Biol. Chem. 265, 9800 (1990).
`8. J. R. Wetterau, K. A. Combs, L. R. McLean, S. N.
`Spinner, L. P. Aggerbeck, Biochemistry 30, 9728
`(1991).
`9. Of the four abetalipoproteinemic subjects, two
`were female (ages 14 and 39 years) and two were
`male (ages 3 and 49 years). All had healthy,
`normolipidemic parents. The subjects had plas-
`ma cholesterol concentrations of between 24 and
`58 mg/dl and no detectable plasma apoB. Their
`current therapy includes low-fat diets and lipid-
`soluble vitamin supplements. The two adult pa-
`tients were the subject of a previously described
`study (20).
`10. TG transfer activity was measured by a modifica-
`tion of a previously described method (4). Before
`performing intestinal biopsies, we explained the
`nature of the study and its possible consequenc-
`es to the subjects or their guardians and obtained
`informed consent. In some instances biopsy sam-
`ples used in this study were part of material
`obtained for diagnostic purposes. Intestinal biop-
`sy samples were frozen and stored at -70°C until
`analyzed. Biopsy tissue was homogenized in a
`polytron homogenizer (Polytron PT3000, Brink-
`mann) at half-maximal setting. Typically, one sam-
`ple was homogenized in 0.25 ml of homogeniza-
`tion buffer [50 mM tris (pH 7.4), 50 mM KCI, 5mM
`
`ing that the presence of intracellular lipid
`droplets does not preclude the measurement
`of TG transfer activity.
`Soluble proteins obtained after deter-
`gent treatment of the intestinal biopsy tis-
`sue homogenates were analyzed by protein
`immunoblotting (13) with antibodies to the
`88-kD component of bovine MTP (14).
`The initial immunoblot analysis of two
`control subjects was performed with anti-
`bodies that had been affinity purified with
`bovine MTP. In both control subjects, a
`single band corresponding to the 88-kD
`component of bovine MTP was identified
`(Fig. 2A). To increase the probability of
`detecting small amounts of the 88-kD com-
`ponent of MTP (15), we used unfraction-
`ated antiserum in subsequent analyses, even
`though some cross-reactivity with other
`
`Table 1. TG transfer activity in intestinal biopsy
`samples.
`
`Subjects
`
`Normalized TG
`transfer activity*
`
`0.33 ± 0.16
`0.011 ± 0.004
`0.28
`0.18
`
`Normal controls (n = 5)
`Abetalipoproteinemia (n = 4)
`Anderson's disease (n = 1)
`Homozygous hypobeta-
`lipoproteinemia (n = 1)
`Nonfasted control (n = 1)
`0.36
`*The activity of a bovine MTP standard was measured
`each time an assay was performed. The TG transfer
`activity from each intestinal biopsy sample was divid-
`ed by that of the standard MTP to normalize the
`activities between experiments. For the first two groups
`of subjects, TG transfer activity is the mean ± SD.
`
`proteins was apparent. Bands comparable to
`that of the 88-kD component of bovine
`MTP were observed in all six control sub-
`jects examined (Fig. 2, B and C). In con-
`trast, no protein corresponding to the 88-
`kD component was detected in the four
`abetalipoproteinemic subjects (Fig. 2D). A
`similar analysis was performed with the
`unfractionated intestinal biopsy tissue ho-
`mogenates from two of the abetalipopro-
`teinemic subjects. Again, no band corre-
`sponding to the 88-kD component of MTP
`was apparent. Two bands with mobilities
`intermediate between the 58- and 88-kD
`components of MTP were present in all six
`control and four abetalipoproteinemic sub-
`jects examined with unfractionated antise-
`rum. Because these bands were not ob-
`served with affinity-purified antibodies (Fig.
`2A), they have been attributed to contam-
`inating antibodies that are specific for pro-
`teins other than MTP. As a control, immu-
`noblot analysis with antibodies to PDI
`demonstrated the presence of the 58-kD
`component of MTP (PDI) in the two abe-
`talipoproteinemic subjects tested (16).
`Our study suggests that MTP plays an
`obligatory role in the assembly of VLDL in
`the liver and chylomicrons in the intestine,
`probably by mediating the transport of lipid
`molecules from their site of synthesis in the
`endoplasmic reticulum (ER) membrane to
`nascent lipoprotein particles within the ER
`as they are assembled. This model for lipo-
`protein assembly is consistent with previous
`studies: Higgins and Hutson (17) showed
`
`A
`
`1
`
`2
`
`3
`
`B
`
`1
`
`2
`
`3
`
`4
`
`88-
`
`88-
`
`58-
`
`Fig. 2. Immunoblot analysis of MTP. Aliquots of
`purified bovine MTP (lane 1 of all four panels) or
`the 103,000g supernatant after treatment of
`homogenized intestinal biopsy samples with
`deoxycholate were subjected to SDS-poly-
`acrylamide gel electrophoresis and immuno-
`blotting with either antiserum or affinity-purified
`antibodies to the 88-kD subunit of MTP. (A)
`Lanes 2 and 3, soluble protein corresponding
`to 34 and 25 ±g
`of homogenate protein, re-
`spectively, from two normal subjects, probed
`with affinity-purified antibodies. (B) Lanes 2 to
`4, soluble protein corresponding to 23 >g of
`homogenate protein from three additional nor-
`mal subjects, probed with unfractionated anti-
`serum. (C) Lane 2, soluble protein correspond-
`ing to 15 pg
`of homogenate protein from a
`subject with Anderson's disease; lane 3, solu-
`ble protein corresponding to 25 ,ug of homog-
`enate protein from an individual with homozy-
`gous hypobetalipoproteinemia; lane 4, soluble
`protein corresponding to 25 pgg of homogenate
`protein from a nonfasted normal individual. Sam-
`ples were probed with unfractionated antiserum.
`(D) Lanes 2 to 5, soluble protein corresponding
`to 18 ,g (lane 2) and 23 ig (lanes 3 to 5) of
`homogenate protein from four abetalipoproteinemic subjects, probed with unfractionated antiserum.
`Lanes 6 and 7, 100 1g of unfractionated intestinal homogenate protein (from abetalipoproteinemic
`subjects corresponding to lanes 4 and 5) were subjected to electrophoresis and immunoblotting with
`unfractionated antiserum. The mobilities of the 58- and 88-kD components of bovine MTP are
`indicated. The figure represents a composite from several independent immunoblots.
`6 NOVEMBER 1992
`SCIENCE * VOL. 258
`*
`
`c
`
`1
`
`2
`
`3
`
`4
`
`D
`
`1
`
`2
`
`3
`
`4 5
`
`6 7
`
`88-
`
`58
`
`88-
`
`58-
`
`-
`
`1 000
`
`2 of 4
`
`PENN EX. 2056
`CFAD V. UPENN
`IPR2015-01835
`
`
`
` on June 6, 2016
`
`http://science.sciencemag.org/
`
`Downloaded from
`
`.,1,.,,,.,,..mp
`
`EDTA, leupeptin (5 jig/ml), and 2 mM phenyl-
`methylsulfonyl fluoride]. A portion of the homoge-
`nate was diluted to 0.6 ml and adjusted to 1.4%
`SDS before measurement of the protein concen-
`tration [0. H. Lowry, N. J. Rosebrough, A. L. Farr,
`R. J. Randall, J. Biol. Chem. 193,265(1951)]. The
`homogenate was then diluted with homogeniza-
`tion buffer to a protein concentration of -1.75
`mg/mI. One part deoxycholate solution (0.56%,
`pH 7.5) was added to 10 parts diluted homoge-
`nate while mixing. Each sample was incubated at
`4,C for 30 min and then centrifuged at 103,000g
`for 60 min. The supernatant was removed, diluted
`1:1 with 15/40 buffer [15 mm tris (pH 7.4), 40 mM
`NaCI, 1 mM EDTA, and 0.02% NaN3], and then
`dialyzed overnight against 15/40 buffer at 4CC.
`Portions of the dialyzed supematant were as-
`sayed for TG transfer activity, and immunoblot
`analysis was performed to detect the 88-kD MTP
`subunit. TG transfer activity was measured as the
`protein-stimulated rate of TG transfer from donor
`small unilamellar vesicles (SUVs) to acceptor
`SUVs. Vesicles of the desired composition were
`prepared by bath sonication in 15/40 buffer as
`described previously (5). The donor and accep-
`tor PC was labeled by adding traces of [3H]di-
`L-alpha-di-
`palmitoyl PC {phosphatidylcholine
`palmitoyl-[2-palmitoyl-9,10,3H(N)]; 33 Ci/mmol;
`DuPont Biotechnology Systems} to a specific ac-
`tivity of -100 cpm/nmol. Donor vesicles contain-
`ing 40 nmol of egg PC, 0.2 mole percent [14C]tri-
`olein {triolein [carboxyl-14C]; -110 Ci/mol; Du-
`Pont Biotechnology Systems}, and 7.3 mole per-
`cent bovine heart cardiolipin (Sigma) were mixed
`with acceptor vesicles containing 240 nmol of egg
`PC and 0.2 mol percent unlabeled TG, 5 mg of
`fatty acid-free bovine serum albumin, and a por-
`tion of the MTP samples in 0.9 ml of 15/40 buffer,
`and the mixture was incubated for 1 hour at 37°C.
`The transfer reaction was terminated by the addi-
`tion of 0.5 ml of a DEAE-cellulose suspension (5)
`and low-speed centrifugation to sediment selec-
`tively the donor vesicles containing the negatively
`charged cardiolipin. The measured amounts of
`[14C]TG (transferred from donor to acceptor
`SUVs) and [3H]PC (marker of acceptor SUV re-
`covery) were used to calculate the percentage TG
`transfer from donor to acceptor SUVs. First-order
`kinetics were used to calculate the total TG trans-
`fer (5). To calculate the protein-stimulated rate of
`TG transfer, the rate of TG transfer in the absence
`of transfer protein was subtracted from that in the
`presence of MTP. To confirm that TG hydrolysis
`was not interfering with our ability to measure lipid
`transfer, after the assay of two subjects we ex-
`tracted the acceptor vesicle lipid (which con-
`tained the transported lipid) and confirmed the
`identity of the TG by thin-layer chromatography.
`All the 14C had a mobility identical to that of
`authentic TG, confirming that intact TG was trans-
`ported in the assays. In addition, the human MTP
`was characterized for its heat stability. MTP was
`inactivated when heated to 60°C for 5 min. The
`loss of activity demonstrates that transfer activity
`attributable to an intracellular form of the cho-
`lesteryl ester transfer protein, which is stable at
`600C [J. lhm, J. L. Ellsworth, B. Chataing, J. A. K.
`Harmony, J. Biol. Chem. 257, 4818 (1982)], was
`not being measured.
`11. The clinical description and other relevant infor-
`mation for patient M.K. with Anderson's disease
`are presented elsewhere [F. Lacaille et al., Arch.
`Fr. Pediatr. 46, 491 (1989); M.-E. Bouma, I. Beu-
`cler, L. P. Aggerbeck, R. Infante, J. Schmitz, J.
`Clin. Invest. 78, 398 (1986)].
`12. The clinical description and relevant data for
`patient C.D. with homozygous hypobetalipopro-
`teinemia are presented elsewhere [G. Gay et al.,
`Rev. Med. Interne 11, 273 (1990); J.-Y. Scoazec
`et al., Gut 33, 414 (1992)].
`13. To identify the 88-kD component of MTP in tissue
`homogenates, we fractionated aliquots of proteins
`to be tested by SDS-polyacrylamide gel electro-
`phoresis and then transferred the separated pro-
`teins to nitrocellulose with a Bio-Rad Trans-blot
`cell. After incubation with a nonfat milk solution, the
`
`W-
`
`&,.9.EKWA
`
`1...
`
`-'I', 'M
`
`'10. -
`
`.... -
`
`'ll, 1,1'~
`
`W, "'
`&..
`1.
`
`1-11m
`
`nitrocellulose filter was incubated ovemight at
`room temperature with an aliquot of antiserum to
`the 88-kD protein (1:300 dilution) or affinity-purified
`antibodies (1:25 dilution). Immunoreactive pro-
`teins were visualized with horseradish peroxidase-
`coupled goat antibodies to rabbit immunoglobulin
`G (Bio-Rad) and a standard developing solution.
`14. The production and characterization of the anti-
`serum to the 88-kD protein have been previously
`described (7). The antiserum immunoprecipitates
`MTP protein and activity, but direct inhibition of
`MTP activity has not been demonstrated. Affinity-
`purified antibodies were prepared as follows:
`Purified MTP (8 to 10 mg) was coupled to 4 ml of
`Bio-Rad Affigel 15. Antibodies were partially puri-
`fied from the antiserum by (NH4)2S04 precipita-
`tion [226 mg of (NH4)2S04 per milliliter of serum].
`After centrifugation, the pellet was suspended
`and applied to the MTP-affigel. The column was
`of 10 mM tris (pH 7.5),
`washed with 100 ml
`followed by 100 ml of 10mM tris (pH 7.5) contain-
`ing 500 mM NaCI. Antibodies were eluted with 50
`ml of 100 mM glycine (pH 2.5) into 5 ml of 1 M tris
`(pH 8.0).
`15. With unfractionated antiserum, the 88-kD band of
`MTP was detectable in the soluble proteins re-
`leased from <3>g of intestinal homogenate pro-
`tein of three of four control subjects investigated.
`
`In the fourth subject, who had the lowest level of
`TG transfer activity of the controls, the 88-kD
`component of MTP was detectable in the soluble
`ig of intestinal homogenate
`proteins from 10
`protein.
`16. J. R. Wetterau et at., unpublished data.
`17. J. A. Higgins and J. L. Hutson, J. Lipid Res. 25,
`1295 (1984).
`18. D. R. Janero and M. D. Lane, J. Biol. Chem. 258,
`14496 (1983).
`19. K. Bostrom et al., ibid. 263, 4434 (1988).
`Clin. Invest. 78, 1707
`20. K. J. Lackner et al.,
`J.
`(1986).
`21. We thank F. Maurer Chagrin (H6pital Bichat, Par-
`is, France) for providing biopsy samples from
`normal subjects, N. Verthier for technical assis-
`tance, and the University of Cincinnati for the
`antiserum to the 88-kD component of MTP. Sup-
`ported by the Bristol-Myers Squibb Pharmaceuti-
`cal Research Institute, the Centre National de la
`Recherche Scientifique, the Institut National de la
`Sant6 et Recherche Medicale (grant CRE 910301
`and a France-U.S. cooperation grant), A.R.C.O.L.
`(the Committee for the Coordination of Research
`on Cholesterol in France), and NIH grant HL1 8577
`(to L.P.A.).
`
`26 May 1992; accepted 25 August 1992
`
`Tyrosine Phosphorylation of CD22
`During B Cell Activation
`Roberta J. Schulte, Mary-Ann Campbell, Wolfgang H. Fischer,
`Bartholomew M. Sefton*
`Ugation of the antigen receptor on B cells induces the rapid phosphorylation of tyrosine
`on a number of cellular proteins. A monoclonal antibody that recognized a tyrosine-
`phosphorylated cell surface protein that was present in activated B cells was generated.
`Amino acid sequence analysis showed that this 140-kilodalton protein was CD22, a B
`cell-specific cell surface glycoprotein and putative extracellular ligand of the protein ty-
`rosine phosphatase CD45. Tyrosine phosphorylation of CD22 may be important in B cell
`signal transduction, possibly through regulation of the adhesiveness of activated B cells.
`
`The B lymphocyte antigen receptor com-
`plex consists of membrane immunoglobulin
`(Ig), at least two accessory molecules (Ig-a
`and Ig-0) (1), several members of the Src
`family of protein tyrosine kinases (2, 3),
`and a 72-kD protein tyrosine kinase that
`may be encoded by the syk gene (4, 5).
`Cross-linking of surface Ig induces rapid
`increases in both tyrosine protein phos-
`phorylation (6-8) and inositol phospholip-
`id hydrolysis (9). Evidence suggests that the
`increased inositol phospholipid hydrolysis is
`induced, at least in part, by tyrosine phos-
`phorylation. (i) Phospholipase C-y, which
`is regulated by tyrosine phosphorylation in
`fibroblasts (10), is phosphorylated on tyro-
`sine during B cell activation (1 1). (ii) The
`increase in free intracellular Ca2+ that re-
`sults from inositol phospholipid hydrolysis
`
`R. J. Schulte, M.-A. Campbell, B. M. Sefton, Molecular
`Biology and Virology Laboratory, The Salk Institute,
`San Diego, CA 92186.
`W. H. Fischer, Peptide Biology Laboratory, The Salk
`Institute, San Diego, CA 92186.
`*To whom correspondence should be addressed.
`
`SCIENCE * VOL. 258
`
`*
`
`6 NOVEMBER 1992
`
`is prevented by treatment of B cells with
`herbimycin, an inhibitor of tyrosine protein
`phosphorylation (12). (iii) Expression of
`the protein tyrosine phosphatase CD45 is
`required for the stimulation of phosphati-
`dylinositol hydrolysis in a murine plasmacy-
`toma (13).
`Protein tyrosine phosphorylation may in
`fact represent the trigger or initial intracel-
`lular biochemical signaling event induced
`by the ligation of surface lg. It is not clear
`how ligation of this receptor complex in-
`duces increased substrate phosphorylation,
`but it is likely that the Src-family kinases or
`the 72-kD kinase plays a role.
`Phospholipase C-y is not the only pro-
`tein to undergo rapid tyrosine phosphoryla-
`tion after cross-linking of surface Ig with
`antibody. Approximately ten newly phos-
`phorylated proteins can be detected by im-
`munoblotting of total cell lysates with an-
`tibodies to phosphotyrosine (6-8), includ-
`ing the vav proto-oncogene product (14),
`the 72-kD cytosolic protein tyrosine kinase
`(4, 6), the 42-kD mitogen-activated (MAP)/
`1001
`
`3 of 4
`
`PENN EX. 2056
`CFAD V. UPENN
`IPR2015-01835
`
`
`
`Absence of microsomal triglyceride transfer protein in individuals
`with abetalipoproteinemia
`Wetterau JR, LP Aggerbeck, ME Bouma, C Eisenberg, A Munck, M
`Hermier, J Schmitz, G Gay, DJ Rader and RE Gregg (November 6,
`1992)
`Science(cid:160)
`
` (5084), 999-1001. [doi: 10.1126/science.1439810]
`
`258
`
` on June 6, 2016
`
`http://science.sciencemag.org/
`
`Downloaded from
`
`(cid:160)E
`
`ditor's Summary
`
`This copy is for your personal, non-commercial use only.
`
`Article Tools
`
`Visit the online version of this article to access the personalization and article
`tools:
`http://science.sciencemag.org/content/258/5084/999
`
`Permissions
`
`Obtain information about reproducing this article:
`http://www.sciencemag.org/about/permissions.dtl
`
`(print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in
`Science
`December, by the American Association for the Advancement of Science, 1200 New York Avenue NW,
`Washington, DC 20005. Copyright 2016 by the American Association for the Advancement of Science;
` is a registered trademark of AAAS.
`all rights reserved. The title
`Science
`
`4 of 4
`
`PENN EX. 2056
`CFAD V. UPENN
`IPR2015-01835
`
`(cid:160)(cid:160)(cid:160)