**4.1 Uptake experiments with inhibitors**

For uptake experiments studied by means of flow cytometry, DNA was labeled with Cy-5. Cell experiments were performed at both 37 °C and 4 °C in order to distinguish cellular uptake from adsorption of lipoplexes to the outside of the cell membrane or their fusion with the cell membrane. It has been described that no energy dependent internalization process can take place at 4 °C [18]. As the size of the DC-30® lipoplexes was determined to be in the range of 300 and 500 nm [5], we assumed internalization to be an active process. The detection of positive fluorescent signal at 4 °C therefore refers to fusion or adsorption of lipoplex with the cell membrane. Fig. 1 represents the amount of fluorescent positive cells after one hour of incubation with lipoplexes at 37 °C.

The corresponding transfection efficiency is listed in Table 2 and shows that despite the fact that uptake of lipoplexes was successful in 24 % of HAEC, less than 1 % of HAEC were transfected. In HASMC, uptake reached as high as 80 %, whereas transfection was only 2 %.

Only 1 % of fluorescent cells were detected after incubation at 4 °C and therefore adsorption to the outside of the cell membrane or fusion with the cell membrane is negligible and the fluorescent positive cells as seen in Fig. 1 can be considered as resulting from cellular uptake (internalization) of the lipoplexes. Analysis of cytotoxicity confirmed results which were published previously [5].

Investigation of Transfection Barriers Involved in

Non-viral cationic lipid-mediated DNAdelivery

types.

by 80 %.

with that achieved with mbCD.

macropinocytosis.

Non-Viral Nanoparticulate Gene Delivery in Different Cell Lines 83

Transfection method Transfection agent Transfection efficiency [%]

Electroporation DNA 43 ± 7 51 ± 8

Fig. 2 and Fig. 3 show the internalization of DC-30 lipoplexes by HAEC and HASMC in the presence of inhibitors of different uptake pathways. Again the incubation at 4 °C did not show any passive adsorption or fusion of DC-30® lipoplexes with the membrane in both cell

As depicted in Fig. 2 A the uptake of DC-30® lipoplexes by HAEC pretreated with filipin (fil) was reduced by about 40 % to 60 % relative to control values. However, another caveolaeinhibitor, genistein (gen), did not reduce lipoplex uptake significantly. Genistein affects tyrosine kinase [19], which associates with caveolae. In contrast, filipin affects 3-ßhydroxycholesterols [20] and avoids the formation of caveolin coats ab initio. Filipin seems to intervene at the origin of internalization whereas genistein seems to inhibit further

The inhibitor methyl-ß-cyclodextrin (mbCD) is involved in cholesterol depletion of the plasma membrane and therefore influencing both cholesterol-rich domains (caveolae and clathrin-mediated endocytosis) and was shown to reduce lipoplex uptake significantly

Clathrin-dependent endocytosis was analyzed with chlorpromazine (chlp) which interacts with clathrin-coated pits and causes their loss from the surface membrane. Uptake was reduced by about 25 %. Combining the clathrin and caveolae-mediated endocytosis inhibitors chlorpromazine and filipin leads to a comparable reduction of lipoplex uptake

Macropinocytosis is possibly involved in the uptake mechanism of DC-30® lipoplexes in HAEC. Pre-incubation with the macropinocytosis inhibitors LY29004 (Ly) and wortmannin

Nocodazol (noco) acts as an inhibitor by depolymerizing microtubules and therefore prevents the transport vesicle (early endosome) from fusing with the late endosome in order to protect the early endosome´s content from digestion and degradation in the lysosomal

Incubation with cytochalasin D (Cch-D), a disrupter of the actin cytoskeleton, resulted in a reduction of about 10 % and did not seem to be strongly involved in the uptake or processing of DC-30® lipoplexes. In this context, the involvement of macropinocytosis, which is strongly actin dependent, and the different inhibition rates caused by LY and wortmannin does not clarify whether or not lipoplexes are internalized via

Fig. 2 B presents data on the internalization of DC-30® lipoplexes by HASMC. Cells pretreated with the caveolae-blocking reagents filipin and genistein showed only a slightly decreased uptake of lipoplexes. In contrast, upon incubation with the clathrin inhibitor

(wm) only led to a significant reduction of uptake using Ly but not wortmannin.

compartment. Applying this inhibitor showed a 50 % reduction in fluorescent signal.

DC-30 lipoplex + 100µM

DC-30 lipoplex + 100 µM

Table 2. Influence of transfection method on transfection efficiency

processing of caveosomes after budding out of the membrane.

HAEC HASMC

DC-30 lipoplex 0,40 ± 0,02 2,01 ± 0,08

Chloroquine 0,70 ± 0,02 2,74 ± 0,14

Chloroquine 0,11 ± 0,03 2,47 ± 0,22

Fig. 1. Cells were incubated for one hour with Cy-5 labeled lipoplexes and analyzed by means of flow cytometry (Fl-4 channel). Prior to the measurement, every sample was additionally incubated with 7-AAD to determine the amount of dead cells (Fl-3 channel). Error bars indicate standard deviation (n=3 in 3 independent experiments).

A HAEC

0

lipoplex

0

lipoplex

fil

gen

mbCD

chlp

Error bars indicate standard deviation (n=3 in 3 independent experiments).

ly

Fig. 1. Cells were incubated for one hour with Cy-5 labeled lipoplexes and analyzed by means of flow cytometry (Fl-4 channel). Prior to the measurement, every sample was additionally incubated with 7-AAD to determine the amount of dead cells (Fl-3 channel).

wm

uptake 37 °C uptake 4 °C dead cells 37 °C dead cells 4 °C

noco

cch-D

chlp+fil

DNA

ctrl

20

40

60

**% fluorescent positive cells**

80

100

120

fil

gen

mbCD

chlp

ly

wm

uptake 37 °C uptake 4 °C dead cells 37 °C dead cells 4 °C

B HASMC

noco

cch-D

chlp+fil

DNA

ctrl

20

40

60

**% fluorescent positive cells**

80

100

120


Table 2. Influence of transfection method on transfection efficiency

Fig. 2 and Fig. 3 show the internalization of DC-30 lipoplexes by HAEC and HASMC in the presence of inhibitors of different uptake pathways. Again the incubation at 4 °C did not show any passive adsorption or fusion of DC-30® lipoplexes with the membrane in both cell types.

As depicted in Fig. 2 A the uptake of DC-30® lipoplexes by HAEC pretreated with filipin (fil) was reduced by about 40 % to 60 % relative to control values. However, another caveolaeinhibitor, genistein (gen), did not reduce lipoplex uptake significantly. Genistein affects tyrosine kinase [19], which associates with caveolae. In contrast, filipin affects 3-ßhydroxycholesterols [20] and avoids the formation of caveolin coats ab initio. Filipin seems to intervene at the origin of internalization whereas genistein seems to inhibit further processing of caveosomes after budding out of the membrane.

The inhibitor methyl-ß-cyclodextrin (mbCD) is involved in cholesterol depletion of the plasma membrane and therefore influencing both cholesterol-rich domains (caveolae and clathrin-mediated endocytosis) and was shown to reduce lipoplex uptake significantly by 80 %.

Clathrin-dependent endocytosis was analyzed with chlorpromazine (chlp) which interacts with clathrin-coated pits and causes their loss from the surface membrane. Uptake was reduced by about 25 %. Combining the clathrin and caveolae-mediated endocytosis inhibitors chlorpromazine and filipin leads to a comparable reduction of lipoplex uptake with that achieved with mbCD.

Macropinocytosis is possibly involved in the uptake mechanism of DC-30® lipoplexes in HAEC. Pre-incubation with the macropinocytosis inhibitors LY29004 (Ly) and wortmannin (wm) only led to a significant reduction of uptake using Ly but not wortmannin.

Nocodazol (noco) acts as an inhibitor by depolymerizing microtubules and therefore prevents the transport vesicle (early endosome) from fusing with the late endosome in order to protect the early endosome´s content from digestion and degradation in the lysosomal compartment. Applying this inhibitor showed a 50 % reduction in fluorescent signal.

Incubation with cytochalasin D (Cch-D), a disrupter of the actin cytoskeleton, resulted in a reduction of about 10 % and did not seem to be strongly involved in the uptake or processing of DC-30® lipoplexes. In this context, the involvement of macropinocytosis, which is strongly actin dependent, and the different inhibition rates caused by LY and wortmannin does not clarify whether or not lipoplexes are internalized via macropinocytosis.

Fig. 2 B presents data on the internalization of DC-30® lipoplexes by HASMC. Cells pretreated with the caveolae-blocking reagents filipin and genistein showed only a slightly decreased uptake of lipoplexes. In contrast, upon incubation with the clathrin inhibitor

Investigation of Transfection Barriers Involved in

transferrin alexa-fluor 488 Rh-DC-30 lipoplexes

cholera toxin-B alexa-fluor 488

single colour images.

**4.2 Microscopic uptake studies** 

mechanism.

spectral-imaging.

A

B

Non-Viral Nanoparticulate Gene Delivery in Different Cell Lines 85

Rh-DC-30 lipoplexes SUN-picture

Fig. 3. HAEC were incubated for one hour with either (A) clathrin marker tf alexa-fluor 488 and Rh-lipoplexes or (B) caveolae marker choleratoxin-subunit B alexa-fluor 488 and Rhlipoplexes. After subsequent DAPI staining of the nucleus, pictures were taken by spectral bio-imaging. I: localization of Rho-lipoplexes; II: localization of markers; III: overlay of

chlorpromazine the highest reduction of uptake was observed (about 85 %). The same result could be achieved with mbCD, which is involved in both clathrin and caveolae-mediated pathways. Again, when combining the specific caveolae and clathrin inhibitors (chlp and fil), lipoplex uptake could not be further reduced. Macropinocytosis is also supposed to be involved in lipoplex uptake in these cells, but pre-treatment with wortmannin alone caused only a minor reduction (about 10 %) whereas Ly did not seem to effect the uptake

On the other hand, uptake of DC-30® lipoplexes by HASMC pre-treated with nocodazole was diminished by about 45 %, which shows a correlation to the clathrin-blocking reagents. The actin-disrupting agent Cch-D did not cause a reduction of lipoplex uptake at all.

The results obtained with uptake studies were confirmed using spectral-bio-imaging. Cells were first pre-incubated with the inhibitors chlormpromazine or filipin. Therefore, the internalization routes were examined via co-localization studies with markers for the clathrin- or caveolae-dependent pathways. The DNA of the lipoplex was labeled with Rhodamine (Rh). Transferrin alexa-fluor 488 was used as a clathrin marker, Choleratoxinsubunit B-Alexa Fluor 488 was a marker for internalization via caveolae-dependent endocytosis and FITC-dextran a marker for macropinocytosis. During preparation for the microscopic experiments, cells were additionally incubated with DAPI (nucleus staining) as a control. Co-localization of the lipoplex and pathway markers was determined by bio-

In HAEC after one h incubation time, co-localization of lipoplexes with tf-488 (Fig. 3A) as well as with chltx-B-488 (Fig. 3B) was detected which indicates uptake via clathrin- and

10µm

10µm

SUN-picture

Fig. 2. After pre-incubation with different inhibitors (see Table 1) (x-range), HAEC (A) and HASMC (B) were incubated for one hour with Cy-5 labeled DC-30® lipoplexes and analyzed by means of flow cytometry (Fl-4 channel). Prior to measurement, every sample was also incubated with 7-AAD to determine the amount of dead cells (Fl-3 channel). The fluorescence intensity of lipoplexes without inhibitors was set to 100%. Error bars indicate standard deviation (n=3 in 3 independent experiments).

A HAEC

0

lipoplex

0

lipoplex

fil

gen

standard deviation (n=3 in 3 independent experiments).

mbCD

chlp

incubated with 7-AAD to determine the amount of dead cells (Fl-3 channel). The

ly

wm

uptake 37 °C uptake 4 °C dead cells 37 °C dead cells 4 °C

Fig. 2. After pre-incubation with different inhibitors (see Table 1) (x-range), HAEC (A) and HASMC (B) were incubated for one hour with Cy-5 labeled DC-30® lipoplexes and analyzed by means of flow cytometry (Fl-4 channel). Prior to measurement, every sample was also

fluorescence intensity of lipoplexes without inhibitors was set to 100%. Error bars indicate

noco

cch-D

chlp+fil

DNA

ctrl

20

40

60

**% fluorescent positive cells**

80

100

120

fil

gen

mbCD

chlp

ly

wm

uptake 37 °C uptake 4 °C dead cells 37 °C dead cells 4 °C

B HASMC

noco

cch-D

chlp+fil

DNA

ctrl

20

40

60

**% fluorescent positive cells**

80

100

120

Fig. 3. HAEC were incubated for one hour with either (A) clathrin marker tf alexa-fluor 488 and Rh-lipoplexes or (B) caveolae marker choleratoxin-subunit B alexa-fluor 488 and Rhlipoplexes. After subsequent DAPI staining of the nucleus, pictures were taken by spectral bio-imaging. I: localization of Rho-lipoplexes; II: localization of markers; III: overlay of single colour images.

chlorpromazine the highest reduction of uptake was observed (about 85 %). The same result could be achieved with mbCD, which is involved in both clathrin and caveolae-mediated pathways. Again, when combining the specific caveolae and clathrin inhibitors (chlp and fil), lipoplex uptake could not be further reduced. Macropinocytosis is also supposed to be involved in lipoplex uptake in these cells, but pre-treatment with wortmannin alone caused only a minor reduction (about 10 %) whereas Ly did not seem to effect the uptake mechanism.

On the other hand, uptake of DC-30® lipoplexes by HASMC pre-treated with nocodazole was diminished by about 45 %, which shows a correlation to the clathrin-blocking reagents. The actin-disrupting agent Cch-D did not cause a reduction of lipoplex uptake at all.
