Results and discussion
Protein aggregation in the presence of tungstate was observed from pH 4 to 5.5, but not at pH 6 or 7. Aggregation was blocked
by increasing the ionic strength with NaCl. The concentration of salt required to prevent precipitation was protein-dependent,
but also varied with both pH and tungstate concentration. Once formed, precipitates could be resolubilized by raising the
pH to 7. Not all proteins precipitated in the presence of tungstate, including those with low isoelectric points. Unlike glass
prefilled syringes, which contained measurable amounts of tungsten and silicone, syringes made of plastic were both tungsten
Aggregation by tungstate requires acidic pH.
Assays were carried out with bovine IgG (bIgG) over the pH range 4–7 at 1, 10, and 100 ppm tungstate. Figure 1 shows that
the bIgG monomer is quantitatively precipitated by 10 and 100 ppm tungstate at pH 4.0–5.5, but is not affected at pH 6.0 or
7.0. The data also show that lower tungstate (1 ppm) is not as effective as higher concentrations at precipitating bIgG.
Figure 1: Effect of pH on the precipitation of bIgG by tungstate. Soluble monomer was quanititated by size-exclusion chromatography.
(ALL FIGURES ARE COURTESY OF THE AUTHORS)
The effect of ionic strength on protein precipitation by tungstate.
The effect of ionic strength on the precipitation of proteins by tungstate was investigated as many drug formulations contain
salt. Figure 2 shows the effect of ionic strength on the precipitation of solutions of Chymotrypsinogen A (Chy A) and bIgG
by tungstate at pH 5.0. The data indicate that 100 mM NaCl was sufficient to completely block the precipitation of ChyA by
tungstate. In contrast, more than 500 mM NaCl was required to completely prevent precipitation of bIgG under the same conditions.
In addition, higher concentrations of salt were necessary to block aggregation in the presence of 100 ppm tungstate than 10
ppm tungstate, and higher ionic strength was required to block aggregation by tungstate at pH 5.0 than at pH 5.5 (data not
shown). These experiments show that the interaction of tungstate with proteins is primarily electrostatic and that the interaction
is stronger at lower pH.
Figure 2: Effect of ionic strength on the precipitation of proteins by tungstate. Soluble protein was quantitated by size-exclusion–high-performance
Using turbidity to evaluate protein aggregation by tungstate.
Measurement of protein turbidity caused by tungstate aggregation provides a rapid alternative to SEC. Turbidity was monitored
spectrophotometrically by measuring the change in absorbance at 350 nm, a wavelength at which protein absorbance is negligible.
Figure 3 compares data obtained for four proteins using both methods. The concentrations required to precipitate one-half
of the protein (IC50s) for both methods are 10–20 ppm tungstate. An assay based on turbidity could also be performed in a
microplate format which would increase throughput and reduce sample handling.
Figure 3: Comparison of protein aggregation by size-exclusion–high-performance liquid chromatography (SEC) with turbidity.
Aggregation by tungstate is correlated with protein net charge.
The interaction of tungstate polyoxyanions appears to be primarily electrostatic; therefore, proteins with a net positive
charge at a particular pH should be precipitable at that pH. When tested at pH 4, there was little difference between the
IC50s for bovine serum albumin (BSA, isoelectric point (pI) 4.2–4.9), ovalbumin (pI 4.5–4.7), and the basic proteins of ChyA
(pI 8.8–9.6) and lysozyme (pI 10.5–11) (see Figure 4). These four proteins all differ in mass, isoelectric point, net charge,
and total number of positive charges (see Table I). However, pepsin, which has a pI below 3 and a net negative charge at pH
4, was not precipitated by tungstate even at 1000 ppm (data not shown). This result is consistent with the proposal that the
net charge on a protein is a key determinant of precipitability.
Figure 4: Effect of pH on the precipitation of ovalbumin and lysozyme. Turbidity was measured spectrophotometrically.
Based on these observations, the authors tested whether the pH range over which a protein would be precipitated by tungstate
could be predicted. The pH dependence of ovalbumin (pI 4.5–4.7) was compared with lysozyme (pI 10.5–11.0). Figure 4 shows
that lysozyme could be precipitated from pH 4 to 5.5, a range over which it remains positively charged. In contrast, ovalbumin
was only precipitable at pH 4 and 4.5, but at pH 5 (slightly above its isoelectric point and close to where it is electroneutral),
it was no longer precipitable.
The interaction of proteins with tungstate is reversible.
To determine whether protein/tungstate precipitates are reversible, a set of precipitates of BSA and bIgG were prepared using
pH 5.0 buffer and 100 ppm tungstate, conditions shown to quantitatively precipitate each protein. The precipitates were re-suspended
in pH 5.0 or pH 7.0 buffer with or without 0.5 M NaCl. Table II shows that dilution into incubation buffer in the absence
of tungstate did not solubilize the precipitates. The addition of 0.5 M NaCl at low pH solubilized about 11% of bIgG. Re-suspension
in pH 7.0 buffer solubilized greater than 90% of BSA and ~60% of bIgG. However, inclusion of 0.5 M NaCl in pH 7.0 buffer completely
Table II: Reversibility of protein aggregates
Tungsten- and silicone-free prefilled syringes for sensitive proteins.
Unlike staked needle syringes made of glass, manufacture of the specialty plastic syringe system (CZ, Daikyo) does not utilize
a tungsten pin to create an opening for inserting the needle. Consequently, this syringe system is expected to be tungsten-free.
Analysis showed that the amount of extractable tungsten in the plastic syringes is at the level of the solvent blank whereas
the amount of tungsten in syringes made of glass is significant (see Table III).
Table III: Tungsten and silicone content of prefilled syringes
Silicone is present in glass syringes as a lubricant. However, silicone has been observed to cause changes in native protein
structure as well as aggregation (7, 8). Analysis showed that specialty plastic syringes, which do not contain silicone oil
as a lubricant, are silicon free (see Table III).