Vol. 91, Issue 1, 211-217, July 2001
Further analysis of transcytosis of free polypeptides and
polypeptide-coated nanobeads in rabbit nasal mucosa
Dario
Cremaschi1,
Silvia
Dossena1,
Cristina
Porta1,
Vilma
Rossi2, and
Mario
Pinza2
1 Dipartimento di Fisiologia e Biochimica Generali, Sezione
di Fisiologia Generale, Università degli Studi di Milano,
I-20133 Milano; and 2 ACRAF, Angelini Ricerche, I-00040 S. Palomba Pomezia, Roma, Italy
 |
ABSTRACT |
In rabbit
nasal mucosa, free polypeptides and polypeptide-coated nanospheres are
actively absorbed by the M cells present in specialized areas of the
epithelium. Because polypeptide-coated nanosphere transport was
abolished in the presence of free polypeptides, free polypeptides and
polypeptide-coated nanospheres are shown here to compete. Fluxes of
polypeptide-coated nanospheres with 356, 490, and 548 nm diameters have
been compared. BSA-coated beads were poorly transported, at the same
rate, when bead diameters were 356 or 490 nm [net flux of
~2-2.5 × 106 nanospheres
(nan) · cm
2 · h1];
however, their net transport largely increased toward a value of
25 × 106
nan · cm2 · h1 at a
diameter of 548 nm. Insulin-coated beads displayed a net flux that was
significantly higher than BSA-coated beads but equally were transported
at the same rate (net flux of ~8.0 × 106
nan · cm2 · h1) at
diameters of 356 or 490 nm; once again, their net flux significantly increased toward a value of 25 × 106
nan · cm2 · h1, if the bead
diameter was 548 nm. Insulin plus anti-insulin IgG-coated 490-nm-diameter beads displayed a very high net flux, although not yet
saturating (~60 × 106
nan · cm2 · h1); however, a
significantly lower saturated net flux (once again ~25 × 106
nan · cm2 · h1) was shown
with 548-nm-diameter beads. In conclusion, 1) in the range
of 356-490 nm diameter, net transport was independent of bead
diameter and, conversely, largely dependent on the coating polypeptides, and 2) at 548 nm diameter, nanospheres tended
to be transferred at similar rates independently of coating kind and
the maximal net transport capacity of the mucosa was reduced. The
suspension viscosity largely increased with 548-nm polypeptide-coated nanospheres; this fact is hypothetically proposed to be the cause of
these events.
polypeptide active transport; antigen sampling; M cells; insulin; antibodies
| |
INTRODUCTION |
SPECIFIC ACTIVE
ABSORPTION of polypeptides, supported by a selective
transcytosis, has been shown across the rabbit nasal mucosa (3,
4, 6-8); it occurs through the M cells of specialized epithelial areas and is correlated with organized subepithelial lymphoid tissue (5, 11). Polypeptides can be transported either presented free in the bathing solution or adsorbed on latex nanospheres (3, 4, 6-8). Observations with this
latter system have shown that net transport is selective for
the different polypeptides; antibodies are preferred, especially if
they are oriented with Fc outside and Fab bound
specifically to their antigen adsorbed on the bead (6,
14). The theoretical and actual maximal transport capacities of
the epithelium have been assessed by kinetic analysis of the transport
as a function of bead concentration and coating (14).
The aim of this paper is to further characterize the transcytotic
pathway by investigating 1) whether free polypeptides and polypeptide-coated nanospheres are transported via the same pathway, 2) how nanoparticle size affects fluxes, and 3)
whether the transport rate of the polypeptide carried by nanospheres is
less than, equal to, or larger than the transport rate of the free
polypeptide. Thus the results of this paper, besides the physiological
interest, may have an impact on whether the transport of polypeptides
after nasal application could be increased and how this could be accomplished.
| |
MATERIALS AND METHODS |
Male New Zealand White rabbits (weight of ~3 kg) were
killed by cervical dislocation, and the nasal mucosae from the roof of
both nostrils were mechanically excised, washed with Krebs-Henseleit saline at room temperature, and mounted on frames between two Teflon
chambers, with the area corresponding to the upper concha exposed in
the window (area of 0.3 cm2). The two chambers were filled
with 1 ml of Krebs-Henseleit saline and bubbled with prehumidified 95%
O2 + 5% CO2 to oxygenate the tissue,
maintain the pH at 7.4, and stir the solution.
The experiments were carried out at a monitored temperature of 27 ± 1°C. At this temperature, the ion and polypeptide transport is
one-half of that measured at 37°C, but the isolated mucosa is stable
and viable for hours (6, 9); on the other hand, the
nostril temperature is lower than 37°C. The viability of the epithelium was checked at the beginning, during, and end of each experiment by measuring the transepithelial potential difference (Vms), as previously reported (8,
9); Vms was 4.6 ± 0.1 and 6.6 ± 0.2 mV (submucosa positive) at the beginning and end of the 2-h
experiment, respectively, across the 260 mucosae used.
The unidirectional mucosa-submucosa (Jms) and
submucosa-mucosa (Jsm) fluxes of free insulin or
polypeptide-coated nanospheres were determined for 2 h, one on the
right and the other on the left mucosa from the same animal (with
random combinations), using the previously described protocol (6,
8). The fluxes of free insulin were expressed as picomoles per
centimeter squared per hour
(pmol · cm2 · h1); those of
nanospheres were expressed as nanospheres (nan) per centimeter squared
per hour (nan · cm2 · h1).
Jms and Jsm were
considered unidirectional because the maximal concentration of insulin
or nanospheres reached in the flux chamber was 104 to
105 compared with that applied in the opposite
donor chamber.
Insulin fluxes.
After a 30-min preincubation with Krebs-Henseleit saline, 1 ml of
solution with insulin at the appropriate concentration was added to the
donor chamber. One minute later, 20 µl of saline were withdrawn to
measure the initial insulin concentration; 2 h later, the final
insulin concentration was measured in the total withdrawn saline. With
nominal 6.5 × 106 M insulin, the initial and final
concentrations were 6.2 ± 0.1 and 5.3 ± 0.1 × 106 M (n = 14); with nominal 26 × 106 M insulin, the initial and final concentrations were
24.9 ± 0.3 and 24.0 ± 0.4 × 106 M
(n = 14); with nominal 52 × 106 M
insulin, the initial and final concentrations were 49.2 ± 0.7 and
51.8 ± 0.9 × 106 M (n = 14).
The 2-h transport was referred to the mean insulin concentration in the
donor saline and corrected by referring to the corresponding nominal
concentration (measured flux × nominal concentration/mean
concentration). The total fluid in the flux chamber at the end of the
experiment was diluted (1:2) with Krebs-Henseleit saline plus 14% BSA
and centrifuged for 30 min at 4°C and 14,700 g in a Z360K
Hermle centrifuge (Maschinenfabrik Berthold Hermle, Gosheim, Germany).
Insulin measurements were performed with the "Enzimun-test Insulin"
kit supplied by Boehringer Mannheim (Milan, Italy). The kit was
designed to measure insulin in serum; thus the insulin in the
Krebs-Henseleit solution was not determined. Insulin was again well
determined if BSA was added to the sample of Krebs-Henseleit saline
(final concentration of 7%). Insulin in Krebs-Henseleit saline tends
to form aggregates, especially if salines are vigorously mixed;
aggregates can be avoided by gently bubbling the gas mixture during the
experimental incubation or gently stirring with a Teflon stick; after
BSA addition, aggregates do not form. Finally, because the kit
was prepared with anti-human insulin IgG and these antibodies only
recognize 35% of bovine insulin (which we have generally used), human
insulin was employed for these experiments.
Nanosphere fluxes.
After a 30-min preincubation with Krebs-Henseleit saline, 1 ml of
solution with polypeptide-coated nanospheres at the appropriate concentration was added to the donor chamber. One minute later, 10 µl
saline were withdrawn to measure the initial nanosphere concentration;
2 h later, the final nanosphere concentration was measured in the
total withdrawn saline. Independently of nanosphere diameter, the
nominal concentration was generally 3.3 × 1011
nan/ml. The measured initial and final concentrations were 3.30 ± 0.01 and 3.15 ± 0.02 × 1011 nan/ml
(n = 232).
The latex nanoparticles were "polybead polystyrene microspheres"
with 490 nm mean diameter, as previously used; nanoparticles with 356 or 548 nm mean diameters were also used. All were supplied by
Polysciences (Warrington, PA). The bead concentration in the donor
saline was measured with a Lambda 5 spectrophotometer (Perkin-Elmer, Norwalk, CT) at a wavelength of 600 nm. To prepare the
calibration straight line, the reference bead concentrations were
calculated on the basis of the solid latex contained in the vial
supplied, the nanosphere diameter, and the polystyrene density (nominal concentrations). These concentrations were also directly measured in a
Burker chamber after suitable dilution. Because of their very low
concentration, the nanospheres transported by the mucosa were measured
solely with the use of a Burker chamber. An Eclipse E600 dark-field
microscope (×600 magnification, Nikon Europe, Badhoevedorp, The
Netherlands) was used for the Burker chamber determinations, taking
advantage of the Tyndall effect. The ratio between the measured and
nominal value of the nanosphere number was 1 if a Burker chamber
contained in a 1.5-mm-thick glass slide was used.
The polypeptides were adsorbed on nanoparticles by using a previously
described method (3, 14). The polypeptide concentration in
the medium used for the adsorption process was generally 6.5 × 106 M. Still unbound surface sites were completely
blocked with BSA. When a double coating of the nanospheres with
antibodies was necessary, a first coating by adsorption with 6.5 × 106 M bovine insulin was performed, after which the
still unbound sites were blocked with BSA (as above); the nanospheres
were then resuspended in Krebs-Henseleit saline containing anti-insulin IgG at a concentration of 6.5 × 108 M (1:100
compared with that of insulin). The suspension was incubated at 4°C
for 16 h on an orbital shaker and spun down at 11,000 rpm (7,260 g) for 10 min; the precipitate was then resuspended in Krebs-Henseleit saline. In both cases (monocoating or double coating), the final nanosphere concentration was generally adjusted to 3.3 × 1011 nan/ml. Variations in insulin and nanosphere
concentrations are reported in RESULTS.
Viscosity measurements.
A 100-µl SGE Perfection syringe (SGE International) was used to
measure the viscosity of the different nanosphere suspensions relative
to water. The length and diameter of the capillary were 51 and 1.58 mm,
respectively; together, the piston and related load weighed 32.2 g. The temperature was fixed at 27 ± 1°C. The nanoparticles
(356, 490, and 548 nm diameters) were coated with 6.5 × 106 M insulin (+BSA) and were at a concentration of
3.3 × 1011 nan/ml; for the 548-nm beads only, the
concentration was also diluted to 2.3 × 1011 nan/ml.
Water or bead suspension (100 µl) was sucked into the vertically
fixed syringe, and the piston (with load) was allowed to run down and
drain the fluid gravimetrically for the entire capillary length. The
draining time was measured with a chronometer able to determine
thousandths of a second. The syringe was washed with ethanol after
every fluid change. The viscosity relative to water (
r)
was measured as the ratio between the draining times of the bead
suspensions and water (1).
Materials and salines.
BSA, murine monoclonal anti-human insulin immunoglobulin G1
(anti-insulin IgG or IgGins), and human sodium insulin
(sodium salt) were supplied by Sigma-Aldrich Chemie (Steinheim,
Germany); bovine sodium insulin (sodium salt) was supplied by
Calbiochem (Lucerne, Switzerland).
The Krebs-Henseleit solution had the following composition (in mM):
142.9 Na+, 5.9 K+, 2.5 Ca2+, 1.2 Mg2+, 127.7 Cl, 24.9 HCO
,
1.2 SO
, 1.2 H2PO
,
and 5.5 glucose, pH 7.4. The polypeptide concentration was molal.
Statistics.
The results are expressed as means ± SE. Student's
t-test for unpaired or, when possible, paired data was used
for the statistical analysis. Interpolating straight lines and curves
were calculated by linear and nonlinear regressions.
| |
RESULTS |
Free insulin fluxes.
Figure 1 shows that the two
unidirectional fluxes (Jms and
Jsm) were not statistically different when donor
chambers contained up to 6.5 × 106 M human insulin.
With increasing insulin concentrations, Jsm increased linearly and Jms was statistically
larger than Jsm, so that a net absorptive flux
(Jnet) was measured, with saturation kinetics.
Jnet data points could be well interpolated by
either a sigmoidal curve or a rectangular hyperbola intercepting the x-axis at 6.4 µM (r2 = 1 in
both cases). Half-maximal concentrations (S0.5,
Km) were 18.0 ± 0.0 µM in the former
case and 17.0 ± 1.4 µM in the latter case. The maximal
Jnet (Jmax) was 6.6 ± 0.0 and 7.9 ± 0.3 pmol · cm2 · h1,
respectively. The Hill number for the sigmoid was 3.4 ± 0.0. The
kinetics of the passive flux (Jsm) followed a
straight line up to at least 52 × 106 M insulin
(r2 = 0.9865): y = (0.12 ± 0.02)x. On this basis, it is possible to
calculate a permeability coefficient of 0.33 nm/s.
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 1.
Transepithelial unidirectional mucosa-submucosa
(Jms) and submucosa-mucosa
(Jsm) and net (Jnet) flux
of insulin as a function of insulin concentration in the donor
solution. Data points represent means ± SE; number of experiments
for each data point = 7. Data points of Jsm
are interpolated by a straight line; those of
Jnet can be interpolated by a sigmoid as well as
by a hyperbola with a threshold at ~6.4 µM (equations and
statistics in the text).
 P < 0.01 compared
with the corresponding Jsm; **P < 0.01 compared with zero.
|
|
Competition between free insulin and insulin-coated nanospheres for
transcytosis.
To obtain Jmax of insulin-coated nanospheres,
beads (490 nm diameter, as previously used) were coated in the presence
of 9.9 × 106 M bovine insulin. Figure
2 shows that Jms
was equal to 13.3 × 106
nan · cm2 · h1 and
statistically larger than Jsm (P < 0.01); Jnet was 10.2 × 106
nan · cm2 · h1. This value
was not statistically different from that previously observed for
Jmax (6). If the experiment was
repeated with donor salines containing both nanospheres (coated in the
same way) and free bovine insulin (20 × 106 M), the
passive bead Jsm was statistically unmodified,
whereas Jms decreased to a value not
statistically different from Jsm so that
Jnet was abolished.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 2.
Unidirectional fluxes (Jms,
Jsm) and Jnet of
insulin-coated nanospheres (diameter = 490 nm) in the absence or
presence of 20 × 106 M insulin in the donor saline.
Insulin concentration for bead coating was 9.9 × 106 M; bead concentration was 3.3 × 1011 nan/ml. Histograms represent means ± SE. Open
bars and solid bars, without (n = 8) and with free
insulin (FI; n = 5) in the donor saline, respectively.
P < 0.01 compared
with the corresponding Jsm; **P < 0.01 compared with zero.
|
|
Effect of bead diameter on nanosphere fluxes and donor saline
viscosity.
Nanospheres with smaller (356 nm) or larger (548 nm) diameters than the
diameter we generally adopted (490 nm) and the 490-nm nanospheres were
coated with 6.5 × 106 M BSA or bovine insulin at
the usual fixed concentration of 3.3 × 1011 nan/ml.
Because bead concentration was equal, the total solid latex in the
saline increased with nanoparticle diameter. This fact can cause
changes in viscosity of the donor saline. Thus r was
measured by comparing the draining times of water and bead suspensions
forced to move along a capillary by equal pressures. Figure
3 shows the results obtained. The
r of bead suspensions was significantly larger than that
of water in any case (P < 0.01). However, the
viscosity of suspensions of 356- or 490-nm beads was larger than that
of water by only 1.50 and 1.59 times, respectively, and these two
values were not significantly different from each other.
Conversely,r of the suspension of 548-nm beads was 3.45 and approximated that of the cytosol. An equal result was obtained with
548-nm beads at a little lower concentration (2.3 × 1011 nan/ml).
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 3.
Viscosity of suspensions of 356-, 490-, and 548-nm beads
relative to water. Relative viscosity (r) is shown on
the right ordinate; nanosphere diameter is shown on the
abscissa. The r was measured as the ratio between
draining times of bead suspensions and water forced to move along a
capillary by equal pressures. Draining time is shown on the
left ordinate. Beads (3.3 × 1011 nan/ml)
were coated with 6.5 × 106 M insulin (+BSA)
following the usual procedure (see MATERIALS AND METHODS);
548-nm-diameter beads were also diluted to 2.3 × 1011
nan/ml. Data points are means ± SE (n = 6 in each
case). Data point concerning 2.3 × 1011 nan/ml
concentration (beads with 548-nm diameter) overlaps with that for
3.3 × 1011 nan/ml concentration. A second batch of
coated beads gave the same results (n = 12 in each
case; not shown in the figure). *P < 0.01 compared
with water viscosity; P < 0.01 compared with 360- and 490-nm bead viscosity.
|
|
The unidirectional fluxes of BSA- or bovine insulin-coated nanospheres
with diameters equal to 356, 490, and 548 nm are reported in Table
1; the corresponding
Jnet values are shown in Fig.
4.
View this table:
[in this window]
[in a new window]
|
Table 1.
Mucosa-submucosa and submucosa-mucosa unidirectional fluxes of
nanospheres (with different diameters) coated with 6.5 × 106 M BSA or insulin
|
|
View larger version (32K):
[in this window]
[in a new window]
|
Fig. 4.
Effect of nanosphere diameter on
Jnet of BSA- or insulin-coated nanospheres. The
corresponding unidirectional fluxes are reported in Table 1. Histograms
represent means ± SE; number of experiments per histogram
(left to right) = 4, 6, 12, 1, 23, 8, 9, and
6. Ratios of net transports compared with the corresponding transport
measured with 356-nm nanospheres are reported below the histograms.
OOP < 0.01 compared with zero.
P < 0.01 compared
with the corresponding transports measured with 356 and 490 nm
nanospheres. **P < 0.01 compared with the transport
measured without MIA and DNP. MIA, monoiodoacetic acid (3 × 103 M); DNP, dinitrophenol (104 M).
|
|
With nanospheres of 356 or 490 nm diameter, Jms,
Jsm, and Jnet displayed
values independent of diameter; conversely, Jms
and Jnet were dependent on coating, being
significantly larger when coated with insulin compared with when coated
with BSA, as already previously observed with 490-nm nanospheres
(6). In two experiments with 356-nm beads coated with
bovine insulin, the presence of free bovine insulin in the bath
(20 × 106 M) abolished the
Jnet (not shown), as shown above for the 490-nm beads.
Conversely, nanospheres with 548 nm diameter, compared with the
nanospheres with a lower diameter, displayed 1) a
significant decrease in Jsm by ~10 times and
2) a significantly large increase in
Jms and Jnet.
Jnet of BSA-coated nanospheres increased from ~2 to 14.5 × 106
nan · cm2 · h1, that is, by
about seven times; Jnet of insulin-coated
nanospheres increased from ~8 to 26.5 × 106
nan · cm2 · h1, that is, by
more than three times. Thus the relative increase was larger with BSA
coating, but the Jnet value reached was
significantly larger with insulin coating (P < 0.05).
To check whether the measured Jnet was active,
the tissue was treated with monoiodoacetate (3 × 103 M) and dinitrophenol (104 M) for the
entire 2-h experimental period, to prevent ATP formation by metabolism.
The Vms dropped from 4.5 ± 0.3 to 0.0 ± 0.0 mV (n = 12), and the bead
Jnet dropped to close to zero without
significant changes in the passive flux Jsm
(results reported in Table 1 and Fig. 4).
Using the reference nanospheres with 490 nm diameter, we previously
observed that a very high Jnet, 8 and 30 times
larger than that obtained with insulin and BSA coatings, respectively, was obtained with a double coating of insulin (adsorbed on the nanosphere) and anti-insulin IgG bound to insulin with the
Fc oriented outward (14). This is confirmed in
Table 2 (which reports
Jms and Jsm) and Fig.
5 (which reports
Jnet). On the basis of kinetics measurements,
this Jnet was about one-half of the maximal
theoretical Jnet achievable, when the
concentration of beads with the same coating was increased
(14). With the same double coating but with 548-nm
nanoparticles, not only is no stimulation observed (unlike what happens
for BSA or insulin coating), but a drastic Jnet
decrease actually occurs: Jnet drops
significantly (P < 0.01) from ~59 to 22 × 106
nan · cm2 · h1, a value not
significantly different from that obtained with insulin coating on
these nanospheres. Moreover, this Jnet is at saturation rate: a decrease in nanoparticle concentration from 3.3 to
2.3 × 1011 nan/ml does not significantly change
Jnet, which remains at ~25 × 106
nan · cm2 · h1 (Table 2,
Fig. 5), Conversely, the same reduction in concentration with 490-nm
nanoparticles produces a significant Jnet
decrease from ~59 to 43 × 106
nan · cm2 · h1 (Table 2,
Fig. 5), as would be expected (14). In conclusion, a
Jnet approximately equal to 25 × 106
nan · cm2 · h1, namely,
that already obtained with insulin coating, seems to be the
Jmax achievable with 548-nm beads.
View this table:
[in this window]
[in a new window]
|
Table 2.
Jms and Jsm of nanospheres (with different
diameters) coated with 6.5 × 106 M
insulin + 6.5 × 108 M anti-insulin IgG
|
|
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 5.
Effect of nanosphere diameter and concentration on
Jnet of beads coated with insulin + IgGins. The corresponding unidirectional fluxes are
reported in Table 2. Histograms represent means ± SE; number of
experiments per histogram (left to right) = 22, 12, 12, and 8. Abscissa shows nanosphere concentration in the donor
solution. OOP < 0.01 compared with zero.
*P < 0.05 compared with transport measured with
3.3 × 1011 nan/ml in the donor saline.
|
|
| |
DISCUSSION |
Free insulin passive flux.
We have previously analyzed the passive fluxes of many organic
molecular markers, with different molecular weights, across the
mucosa of the upper concha of the rabbit, including two
polypeptides, namely elcatonin [molecular weight
(Mr) = 3,362] and adrenocorticotropic hormone (Mr = 4,500) (8). We
have shown that, in this mucosa, two populations of inter(para)cellular
pores exist, the first comprising numerous pores, which allow free
diffusion of molecules with Mr up to 1,000, and
the second comprising a few large pores (up to 4-nm radius) allowing
free diffusion (with low permeability) of molecules with
Mr up to 4,000, whose Jsm
kinetics were linear at least up to 25 mM (8). Elcatonin,
which is dimeric in solution and thus has an actual
Mr of 6,724, displayed lower permeability than
these latter markers; kinetics of the passive flux
Jsm saturated over a concentration of only 2.4 µM, indicating movements by drastically restricted diffusion
(4, 8).
The human sodium insulin we have now tested has intermediate
diffusional properties. With a Mr = 5,749.5, i.e., only slightly smaller than that of dimeric elcatonin,
and permeability of the same order as elcatonin, it displays linear
Jsm kinetics at least up to 52 µM. These
observations suggest that insulin can move either at the borderline of
free diffusion or, more probably, with only a less restricted diffusion
than elcatonin dimer.
In any event, the small population of large pores begins to hamper free
diffusion of molecules with Mr = 5,000-7,000 and a few nanometer diameter. Thus it is surprising to
observe passive fluxes of nanospheres with up to ~500 nm diameter,
even if their diffusion is restricted (6, 14). Clearly,
they must move through a third smallest population of pores with 100-nm
diameter but negligible total area, even compared with the sole second pore population. In agreement, the permeability coefficient of the
490-nm beads, which can be calculated from Table 1, is 0.03 nm/s, i.e.,
11 times lower than that of free insulin. Thus the majority of free
polypeptide molecules moves through the larger area of the first two
populations of pores and a negligible minority moves through the third
population. The third class of pores probably is not located in
intercellular junctions but in a very few nanolesions of the
epithelium, such as damaged cells (3, 6).
Free insulin Jnet.
A net absorption of human insulin occurs above a concentration of 6.5 µM. It follows either sigmoidal kinetics, with three cooperating
binding sites, or hyperbolic kinetics, with a threshold at ~6.5 µM
insulin. In any event, Jmax is ~7
pmol · cm2 · h1, about
twice that of elcatonin, and the half-maximal concentration (S0.5, Km) is ~18 µM, more than
five times that of elcatonin (4). This means that the
transcytotic system displays less affinity but a higher transport
capacity for human insulin than for elcatonin. The difference in net
transport capacity (if it is not due to the different assays used or
the different time periods of the measurements) may suggest that
different pathways (receptors?) are involved in the transcytosis of the
two polypeptides. Conversely, insulin and elcatonin adsorbed on
nanospheres display approximately the same Jmax
and the same half-maximal concentration (6). The different
behavior for insulin might also be explained by the fact that bovine
insulin, instead of human insulin, was adsorbed on the nanospheres;
however, a certainly important fact is that the molecules adsorbed can
be forced into positions that can favor or hamper the binding, so as to
change the affinity. Otherwise, it can be that different pathways are
involved in the transport of free polypeptides and polypeptide-coated nanospheres.
On this basis, to clarify whether free insulin and insulin-coated
nanospheres use the same pathway, competition between bovine insulin-coated nanospheres and free bovine insulin, at a concentration certainly larger than Km
(S0.5), was investigated. The result, complete
abolition of bead Jnet, shows that pathways for
either transport are identical.
Effect of nanosphere diameter on fluxes.
The results reported show that BSA- or insulin-coated nanospheres are
transported at a rate independent of nanosphere diameter ranging from
356 to 490 nm. Equally, passive permeability is not significantly
affected. However, the change in diameter from 490 to 548 nm is
critical, as the passive flux Jsm decreases by
~10 times and Jnet increases by 7 times (BSA
coating) or 3 times (insulin coating) and, in absolute values, to a
maximum of ~25 × 106
nan · cm2 · h1 (insulin coating).
The decrease in Jsm could be predicted because
nanospheres with a diameter of 490 nm already cross the mucosa by
restricted diffusion, probably single-file diffusion (6,
14). Thus a further increase in diameter can only reduce permeability.
The increase in Jnet could not be so easily
predicted, since the increase in diameter should reduce the possibility
of bead accommodation in the endocytotic pits and should reduce
transport. Moreover, Jnet increase cannot be due
to an increase in polypeptide loading on larger nanoparticles
because 1) BSA coating corresponds in all kinds of
nanospheres to 100% of bead surface (as sites still unbound after
coating incubation are blocked with BSA), 2) by increasing
insulin loading, 490-nm nanospheres display a maximal transport of only
10 × 106
nan · cm2 · h1, very far
from the required 25 × 106
nan · cm2 · h1 obtained
with 548-nm nanospheres, and this occurs with coating concentrations
very far from those saturating the bead surface (this paper and Ref.
6).
Bead concentration being constant, the change in bead diameter
introduces a change in solution viscosity as the only dependent variable. Whereas viscosity and net transport are not significantly different for 356- and 490-nm beads, viscosity and net transport abruptly increase in parallel for 548-nm nanoparticles. Cytosol viscosity is 5-20 times that of water and, avoiding cells with ectoplasm, nearer to 5 (2, 12, 13). Thus the viscosity of
the suspension of 548-nm beads approaches that of cytosols. It has been
predicted theoretically and shown by experiments on transferrin-mediated endocytosis in two different cell lines that, with
extracellular viscosity close to cytosolic viscosity, the cell plasma
membrane tends to invaginate more rapidly so that endocytic pits and
vesicles accelerate their formation rate; this accelerates endocytosis
(10). At present, we have no direct evidence in favor of a
stimulation of the apical endocytosis in the nasal epithelium under the
conditions described, which may be caused by the observed increase in
viscosity of the bathing saline. Hypothetically, we can suppose that
such a relationship exists so that all apical processes of the
transcytosis and as a consequence the whole pathway are altered. In
this case, the rearrangement must involve a stimulation of
Jnet of BSA- or insulin-coated nanospheres,
which are low with 356- to 490-nm beads; conversely, it must cause a
decrease of the Jnet of insulin + anti-insulin IgG-coated nanoparticles, which is highest with the
smaller beads.
In conclusion, the maximal net transport capacity of the
mucosa is largely reduced to ~25 × 106
nan · cm2 · h1 with the
548-nm beads.
Conclusions.
Because nasal application of polypeptide hormones and related active
compounds is a therapeutic strategy, on the basis of the present
results, it must be concluded that the use of 548-nm beads produces
improvements in the low transports but a drastic decrease in the
maximal transport capacity of the mucosa. With 490-nm beads, which seem
to support the best maximal net transport capacity, the
Jmax of insulin adsorbed on nanospheres can be
compared with free insulin Jmax. At the maximal
concentration of soluble bovine insulin (~470 µM), 2.20 × 107 pmol can be loaded on one 490-nm bead; this
corresponds to 83% of the maximal loading (6). The
maximal loading can be reached with the more soluble human insulin
(2.65 × 107 pmol/nan). Jmax
of insulin-coated nanospheres is ~10 × 106
nan · cm2 · h1 and can be
already reached with low loadings; if net transport is stimulated by a
double coating with insulin + IgGins,
Jmax can reach ~90 × 106
nan · cm2 · h1
(14). Thus the Jmax of insulin
adsorbed on nanospheres can be 2.6 pmol · cm2 · h1 in the
first case (monocoating with only insulin) and 23.9 pmol · cm2 · h1 in the
second case (double coating with insulin + IgGins),
the former being much less (37%) and the latter much more (3.4 times) than free insulin Jmax (~7
pmol · cm2 · h1).
| |
ACKNOWLEDGEMENTS |
This research received financial support from ACRAF Angelini
Ricerche, S. Palomba Pomezia, Roma, Italy.
| |
FOOTNOTES |
Address for reprint requests and other correspondence: D. Cremaschi, Dipartimento di Fisiologia e Biochimica Generali,Via Celoria
26, I-20133 Milano, Italy (E-mail:
fistrasp{at}mailserver.unimi.it).
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 27 October 2000; accepted in final form 12 February 2001.
| |
REFERENCES |
1.
Bull, HB.
An Introduction to Physical Biochemistry. Philadelphia, PA: F. A. Davis, 1964.
2.
Burns, VW.
Measurements of viscosity in living cells by a fluorescent method.
Biochem Biophys Res Commun
37:
1008-1014,
1969[ISI][Medline].
3.
Cremaschi, D,
Porta C,
and
Ghirardelli R.
The active transport of polypeptides in the rabbit nasal mucosa is supported by a specific vesicular transport inhibited by cytochalasin D.
Biochim Biophys Acta
1283:
101-105,
1996[Medline].
4.
Cremaschi, D,
Porta C,
and
Ghirardelli R.
Endocytosis of polypeptides in the nasal respiratory mucosa of the rabbit.
News Physiol Sci
12:
219-225,
1997[Abstract/Free Full Text].
5.
Cremaschi, D,
Porta C,
and
Ghirardelli R.
Relationship between polypeptide transcytosis and lymphoid tissue in the rabbit nasal mucosa.
Biochim Biophys Acta
1369:
287-294,
1998[Medline].
6.
Cremaschi, D,
Porta C,
and
Ghirardelli R.
Different kinds of polypeptides and polypeptide-coated nanoparticles are accepted by the selective transcytosis shown in the rabbit nasal mucosa.
Biochim Biophys Acta
1416:
31-38,
1999[Medline].
7.
Cremaschi, D,
Porta C,
Ghirardelli R,
Manzoni C,
and
Caremi I.
Endocytosis inhibitors abolish the active transport of polypeptides in the mucosa of the nasal upper concha of the rabbit.
Biochim Biophys Acta
1280:
27-33,
1996[Medline].
8.
Cremaschi, D,
Rossetti C,
Draghetti MT,
Manzoni C,
and
Aliverti V.
Active transport of polypeptides in the rabbit nasal mucosa: possible role in the sampling of potential antigens.
Pflügers Arch
419:
425-432,
1991[ISI][Medline].
9.
Cremaschi, D,
Rossetti C,
Draghetti MT,
Manzoni C,
Porta C,
and
Aliverti V.
Transepithelial electrophysiological parameters in rabbit respiratory nasal mucosa isolated "in vitro."
Comp Biochem Physiol
99A:
361-364,
1991.
10.
Edwards, DA,
Gooch KJ,
Zhang I,
McKinley GH,
and
Langer R.
The nucleation of receptor-mediated endocytosis.
Proc Natl Acad Sci USA
93:
1786-1791,
1996[Abstract/Free Full Text].
11.
Ghirardelli, R,
Bonasoro F,
Porta C,
and
Cremaschi D.
Identification of particular epithelial areas and cells that transport polypeptide-coated nanoparticles in the nasal respiratory mucosa of the rabbit.
Biochim Biophys Acta
1416:
39-47,
1999[Medline].
12.
Heilbrunn, LV.
An Outline of General Physiology (3rd ed.). Philadelphia, PA: W. B. Saunders, 1952.
13.
Heilbrunn, LV.
The viscosity of protoplasm.
In: Protoplasmatologia. Handbuch der Protoplasmaforschung. Vienna: Springer-Verlag, 1958, vol. 2, p. 1-109.
14.
Porta, C,
Dossena S,
Rossi V,
Pinza M,
and
Cremaschi D.
Rabbit nasal mucosa: nanospheres coated with polypeptides bound to specific anti-polypeptide IgG are better transported than nanospheres coated with polypeptides or IgG alone.
Biochim Biophys Acta
1466:
115-124,
2000[Medline].
15.
Porta, C,
James PS,
Phillips AD,
Savidge TC,
Smith MW,
and
Cremaschi D.
Confocal analysis of fluorescent bead uptake by mouse Peyer's patch follicle-associated cells.
Exp Physiol
77:
929-932,
1992[Abstract].
J APPL PHYSIOL 91(1):211-217
8750-7587/01 $5.00
Copyright © 2001 the American Physiological Society
TOP
ABSTRACT
INTRODUCTION
