Primary amniocytes can be efficiently transformed by adenoviral E1-functions
In order to test if primary amniocytes can be used to establish
permanent, therapeutic protein expressing cell lines, amniocytes were
cotransfected with the E1-expressing plasmid pGS119 containing E1A-,
E1B- and pIX-functions plus the plasmid pGS116 expressing the human
alpha-1 antitrypsin (hAAT) glycoprotein.
The primary amniocytes used here originated from a single
amniocentesis sample. About 20–30 days after transfection of
approximately 1 × 106 cells, numerous transformed cell
colonies appeared on all dishes, and cells on the dishes were expanded
as 10 pools (Z171-1A, B – Z171-5A, B). Non-transformed cells stop
growth already at early passages and in approximately passage 10–15
mainly transformed cells can be detected on the dishes. Since
transfection efficiency of primary cells is less than 1% (data not
shown) the efficiency of transformation is very high, however, because
cells were passaged twice the number of independently transformed
clones could not be determined. Such high frequencies were obtained
without difference in numerous transfections using different cells from
several amniocentesis (data not shown).
Expression of the E1 proteins was analyzed by Western Blot using
protein extracts of different cell pools. Like HEK293 cells used as
positive control, all cell pools express both the E1A- and 21kD-E1B
proteins, although in different amounts and ratios (Fig. 1).
Primary amniocytes show variable morphologies. In
contrast, transformed cells do appear smaller and show more uniform
morphologies. In very early passages the morphologies in different cell
pools are quite similar but change during further cultivation. Most
transformed cells go through mild crisis and decelerate growth but
recover during only few passages.
Transformed amniocytes express and secrete hAAT
As described above, the primary amniocytes were cotransfected with a
hAAT-expressing plasmid. hAAT is a major human serum protein which is
predominantly produced and efficiently secreted from hepatocytes. Thus
we tested for secreted (Fig. 2a) and intracellular (Fig. 2b)
hAAT in different cell pools. Using Western Blot analyses we were able
to detect hAAT expression in 6 out of 10 different cell pools. The
amount of protein loaded per lane for the intracellular and secreted
protein corresponds to 8 × 104 cells and 4 × 103 cells,
respectively. These results suggest that hAAT is very efficiently
secreted from the cells since only small amounts of the protein were
intracellularly located.
Amniocyte cell lines show high and long-lasting hAAT expression
The above results show that both plasmids integrated upon
cotransfection and cells express proteins from both plasmids. The lack
of selective pressure during cell passaging most likely does not
influence expression of E1-functions since their continuous presence is
expected to be crucial to maintain the transformed phenotype of the
cells. Therefore, we tested for the stability of expression of hAAT
during multiple passages.
Six Z171 cell pools showed expression of hAAT in early passages and
were thus further cultivated up to 35–50 passages. In different
passages 7 × 105 cells were plated, the supernatants were collected and the amounts of secreted hAAT were quantitated. In figure 3
the amount of hAAT in 6 different cell pools is depicted and shows that
4 out of 6 cell pools show high and long-lasting hAAT expression up to
6 μg/ml. Only in two cell pools the hAAT
expression drops drastically to almost not detectable expression levels
after 30 passages. Two cell pools (Z171-5A and Z171-5B) stably express
up to 8 pg hAAT per cell and day for more than 50 passages without any
antibiotic selection.
Since each cell pool is derived from numerous different
transformation and integration events and thus contains many
genetically different cell lines, we performed single cell cloning on
cell pools Z171-5A and Z171-5B by limited dilution in 96-well plates.
Thirteen single cell clones originating from Z171-5A and 25 single cell
clones originating from Z171-5B were expanded and tested for hAAT
expression (data not shown). Figure 4
shows long term expression of hAAT of 3 clonal cell lines each
originating from pool Z171-5A and Z171-5B, respectively. Over 60
passages all 6 clonal cell lines express hAAT; in 2 clonal lines
maximum hAAT levels reach 27 and 30 pg/cell/day.
hAAT expressed in amniocyte cell lines is glycosylated and sialylated
Human alpha-1 antitrypsin is a 396- amino acid serum glyoprotein and
contains three carbohydrate side chains N-linked to asparagine
residues. Analyses of the carbohydrate composition revealed two main A-
and B-types oligosaccharide chains (see Fig 5c) in 2:1 ratio [22,23].
Glycosylation of hAAT does not seem to be important for the formation
of a biologically active conformation with elastase and thus for
activity of the protein [24]. However, glycosylation seems to play a crucial role in stability of hAAT in the serum [25,26] and in secretion from hepatocytes [27].
In order to test if amniocyte cell lines produce glycosylated hAAT we treated the cell culture supernatants of Z171-5A (Fig. 5)
and Z171-5B (data not shown) with PNGaseF, followed by hAAT-specific
Western Blotting. Since PNGaseF cleaves between the innermost GlcNAc
and asparagine residue of oligosaccharides from N-linked glycoproteins
(see Fig. 5c)
we expected a shift in molecular weight if PNGaseF can hydrolyze
N-glycan chains from hAAT. For control we used hAAT purified from human
serum. Digestion with increasing amounts of PNGaseF results in a clear
shift in molecular weight (Fig 5a).
Moreover, appearance of three distinct bands clearly suggests the
presence of three glycan chains in the recombinant product. There is no
difference in size and number of different protein bands when compared
to hAAT purified from human serum.
All galactoside residues in plasma derived hAAT are known to be
linked with sialic acid. To test for this feature in hAAT expressed in
amniocyte cell lines, protein secreted into the cell culture medium of
Z171-5A was digested with Neuraminidase followed by hAAT-specific
Western Blotting (Fig. 5b). Comparable to plasma derived hAAT, the hAAT produced in Z171-5A showed a shift in size when treated with Neuraminidase.
These results indicate that hAAT produced in human amniocyte cell lines are fully glycosylated and sialylated.
Answers to all your Biology Questions
