Estado nutricional de los niños de la IEP “20 de Enero” no 70621

(a) binding to ECM substrates

To provide more direct evidence for the possible interactions of CD33 and CD34 antigens and BGPc protein with ECM, we next studied the adherence to purified ECM components. Attachment was examined at 37°C, and the ECM su b strates used in the adhesion assay were fibronectin, fibrinogen, laminin, collagens I, III, IV, IX, X and hyaluronic acid. Prelim inary adhesion assays with CHO transfectants to the ECM substrates showed, as expected from the above results w ith strom al layers, high background binding and no differences in adherence between the CHO, CHO-CD33, CH0-CD34 and CHO-BGPc cells to th e ECM substrates (data not shown).

We next examined the binding of CD34-Fc and CD33-Fc fusion proteins. No specific adherence was observed when the binding of ^^S- labelled CD33-Fc (Fig. 3.8) or CD34-Fc (data not shown) fusion proteins to the ECM substrates was measured. Similar results were obtained when the binding of CD34-Fc and CD33-Fc fusion proteins to ECM components was detected by ELISA using HRP-conjugated Protein A, followed by ABTS substrate (data not shown) or when the binding was monitored by 125i_iabelled Protein A. The last is shown in Fig. 3.9, where the binding of CD34-Fc fusion protein does not exceed th a t of 125i_iabelled Protein A alone to any of the ECM components (data for CD33-Fc fusion protein and I25i-iabelled Protein A not shown). These results suggest th a t CD34-Fc

Fn Ln C o ll Col III C olIV ColIX ColX HA BSA Total 0 1000 2000 3000 c p m

F ig u r e 3.8: B in d in g o f ^Sg.iabelled CD33-Fc fu sio n p r o te in to ECM su b strates

100 |il of 35S-labeIled CD33-Fc fusion proteins (2 500 cpm) was added to the wells precoated with the indicated ECM substrates (Fn-fibronectin; Fg-fibrinogen; Ln- laminin; Col 1- collagens I, III, IV, IX and X, HA-hyaluronic acid and BSA as a negative control). Binding was performed for 2 hours at 37®C. The counts recovered in the adherent layer in comparision to the total counts added (total) are shown. HA ColX Col DC C ollV Col 111 C o ll Ln Fg Fn BSA Protein A CD34-FC — I— 0.0 1.0 —r- 2.0 —r— 3.0 —r— 4.0 —|— 5.0 6.0 % BINDING

F igu re 3.9: B in d in g o f CD34-Fc fu sion protein to ECM com ponents

200 pg of CD34-Fc protein was preincubated with 1.5 pCi of l^^I-labelled Protein A for 1 hour at room temperature and 100 pi of this mixture was added to the wells precoated with the indicated ECM substrates (Fn-fibronectin; Fg-fibrinogen; Ln-laminin; Col I- collagens I, III, IV, IX, X, HA-hyaluronic acid and BSA as a negative control). As a negative control, the binding of Protein A only was monitored. Binding was performed for 2 hours at 37°C. The percentage of the total radioactivity recovered in the adherent layer is shown.

and CD33-Fc fusion proteins do not adhere specifically to fibronectin, fibrinogen, laminin, collagens I, III, IV, IX, X and hyaluronic acid.

(b) binding to bone marrow ECM extract

Campbell et al. (1987) bave shown that bone marrow ECM, isolated by high salt precipitation, followed by guanidine extraction, preferentially promotes the adhesion of granulocytic bone m arrow precursors. To identify the putative lineage-specific adhesion protein they have separated this ECM extract by SDS-PAGE, transferred the proteins to nitrocellulose and probed using ^^Cr-labelled unfractionated m urine m arrow cells. U sing th is assay they have a ttrib u te d th e strong granulocytic cytoadbesion to a 60 kD baemonectin. In order to test whether CD34 and CD33 antigens interact with such bone marrow ECM extract or w ith baem onectin we prepared bone marrow ECM following Campbell's procedure (Campbell et al., 1985), and probed the ECM blots with ^^Cr- labelled CHO-CD33, CHO-CD34, CHO-BGPc cell transfectants and K G l cells, which as shown by FACscan analysis express CD33 and CD34 antigens (data not shown). No specific binding was observed with any of the CHO cell transfectants or KGl cells to the ECM blots. Since a t th a t time we did not have anti-baemonectin sera, we used as a positive control the binding of ^^Cr-labelled murine bone marrow cells, as described in the original identification of baemonectin (Campbell et al., 1987). However no specific binding of ^iCr-labelled m urine bone marrow cells to the ECM blots was observed.

The ECM blots were then probed with CD33-Fc and CD34-Fc fusion proteins, followed by HRP-conj ugated sheep F(ab')2 anti-bum an Ig.

Specific binding to a 110 kD protein and a weaker band a t 100 kD was detected with CD34-Fc fusion protein (Fig. 3.10). This result was obtained on th ree separate occasions using one batch of CD34-Fc proteins.

- kD

\ ^ ^

205

- 1 1 6

66

45

29

F igu re 3.10: B in d in g o f CD34-Fc and CD33-Fc fu sion p ro tein s to rabbit b one m arrow ECM extract

The rabbit bone marrow extract was separated in 7.5% SDS-PAGE (non­ reducing conditions) and transferred to nitrocellulose membrane. The binding of CD34-Fc (20 pg/ml; lane C), CD33-Fc (20 pg/ml, lane B) fusion proteins and PBS, 0.1% C^/v) BSA (lane A) was performed for 1 hour a t room tem perature, followed by HRP-labelled sbeep F(ab')2 anti-bum an Ig

generously provided by Dr. David Simmons, which preparations unfortunately we have not characterised by western blotting. However no specific binding was observed with the CD34-Fc fusion proteins produced in our laboratary using the Fc-constructs, provided also by Dr. D. Simmons.