5.2 A RTICLE II
5.2.1 Contrast agent selection
a
non-invasive biodistribution study, clearly the signal loss, leading to decrease in the anatomical features would not be desirable. However, while the T1 contrast agents enable better anatomical imaging, the concentrations of metal chelates required for significant increase in the contrast ar
typical dose for gadolinium being 0.1-0.3 mmol/kg total body weight (Brücher and Sherry, 2001). This would mean >30 µmol for 333 g rat and with 10 µl injection of 1*1011 viral vector with 100 biotin binding sites / virus would result in 0.166 pM solution with one Gd3+ per biotinylated chelate. When increasing the number of Gd3+ per chelate by adding chelate structures to a peptide backbone, combining metal to macromolecules (Caravan et al., 1999), or taking in consideration the concentration
8 fold difference between the calculated threshold and reality. The recently published results on Cowpea chlorotic mottle virus with 180 Gd3+ resulted in relaxivity value of 250 mM-1S-1 (Allen et al., 2005) and with nanoparticles 598 mM-1nS-1(Winter et al., 2003). It remains to be seen if these agents can be used in context with gene therapy viruses.
T2 agents have already been shown to be feasible in vivo imaging of even single cells (Shapiro et al., 2006), although in practise more cells, e.g. 40 stem cells (Hoehn et al., 2002), are needed for imaging trafficking and targeting (Arbab et al., 2004). The limits of detection for T2*
weighted imaging in brain have reported to be 2.4 µg Fe/ml with 120*103 cells/ ml (10 µM) (Dahnke and Schaeffter, 2005), significantly smaller than with T1 agents.
While iron ions are part of Fenton’s reaction, catalyzing hydroxyl radicals, it has been shown that ferritin is not toxic in vitro (Genove et al., 2005) and might even offer resistance to chemically induced oxidative stress. However some evidence of toxicity was seen in another study (Muldoon et al., 2005). We performed an
Table 12. MTT assay of the cytotoxicity of the bUSPIO particles. The control cells were calculated as 100%
survival and the results are respectively compared to the control cells, presented as mean ± SEM.
Formulation Survival
Baavi 103 ± 2.1
Wild-type virus 109 ± 3.9
bUSPIO 108 ± 3.6
Baavi + bUSPIO 99 ± 4.3 Wild-type virus + 99 ± 3.9 bUSPIO
Transduction assays in HepG2 cell line showed that both transduction efficiency and expression levels of the beta-galactosidase were higher with bUSPIO-coated Baavi as compared to non-coated Baavi, (II/ fig 2). It could be possible that while the bUSPIO-Baavi complex was too small to sediment on its own, the Brownian motion was somewhat decreased, resulting in increased transduction efficiency as compared to the free virus, comparable to the mechanism seen to increase transduction efficiency of lentiviruses (Chan et al., 2005).
The levels of beta-galactosidase enzyme expression were much higher with the coated viruses as compared to normal Baavi, possibly due to the fact that one USPIO contains on average of 1-2 attached virions, thus resulting in an increase in the amount of capsids transported to the nucleus as complexes entered the cell, increasing the moi per individual cell.
The effect of magnetic targeting in vitro with bUSPIO was compared to the 1.1 µm SPIO particles and showed that due to the small size of the bUSPIO, the effect was less profound with similar conditions (data not shown), as could be expected based on the Equation 1. While observing the beta-galactosidase enzyme levels with both particle types with magnetic targeting, it could be seen that the overall transgene levels were lower with SPIO and also with USPIO, due to concentration of the viruses to a smaller area with less permissive cells (data not shown). This could indicate that while the increase of capsids in cellular entry with a relatively unlimited number of 5.2.2 Atomic force microscopy
The ratio of viruses to the iron particle is crucial when considering the similar size of the viruses and biotinylated USPIO. With previously used SPIO with a mean diameter of 1.1 µm, the particle is five fold larger, therefpre likely to result in a particle coated with viruses. With smaller USPIO with a mean diameter of 50 nm, the ratio and resulting composition was analyzed by using atomic force microscopy. Atomic force microscopy has the ability to yield images under ambient conditions or in a solution, which provides an ideal tool for the physical study of biological specimens under physiological conditions (Horber and Miles, 2003).
It was observed that keeping the viral particles highly saturated (100:1) with multibiotinylated USPIO during the overnight coating, caused limited formation of crosslinked aggregates, resulting in small-sized iron-virus complexes with 1-2 virus per bUSPIO (II/ fig 1).
Together with the 50 nm size of the USPIO as compared to the 25x200 nm virus these factors increased bUSPIO-Baavi diffusion in liquid in vivo, probably resulting in more natural biodistribution and kinetics of the virus.
5.2.3 In vitro transduction
cells increases the transgene expression levels, the limits will be reached with magnetic sgene expression levels.
V) injection of bUSPIO-coated Baavi a specific loss of MRI n ventricle (II/ fig 3). The signal loss remained fig 4). Since baculoviruses are known to exhibit choroid plexus (CP) (Lehtolainen et al., 2002) ight explain the MRI contrast changes in the corresponding choroid plexus cells are known to have extensive al., 2005), the contrast changes could not be explained by passive MRI signal changes was detected in animals diffused to the rat ventricular system (~300 µl) to a . In contrast to other studies, the amount of administered iron en considering the fate of the Baavi-bUSPIO particles not entered, the cerebrospinal uid (CSF) clearance is reportedly rapid (Nagaraja et al., 2005) with one hour turnover in rats
avson and Segal, 1996), and after two hours unbound bUSPIO or non-endocytosed bUSPIO-ped from the CSF (Muldoon et al., 2004).
on of the iron amount in rat brain using MRI (Kroll et al., 1996). It is therefore impossible to estimate the amount of transduced cells based on the iron related signal loss. In this respect, methods based on SPECT or PET are more quantitative (Schellingerhout and Bogdanov, Jr., 2002). However as MRI methods have been reported to reach a spatial resolution of 30 µm (Lee et al., 2001), this accuracy might enable detection of individual cells and provide more vital information after intracerebral viral administration in the future. However, the blooming effect with iron would harden the separation of individual cells from clusters of labelled cells.
concentration, resulting in lower tran
5.2.4 Detection of viral particles by MRI After the intracerebroventricular (IC
signal was found on the injected side of the rat brai detectable for at least two weeks (14 days) (II / strong tropism towards cuboid epithelial cells in the the accumulation of iron to these cells m
area of the rat brain. Even though the secretive endocytotic traffic (Emerich et
endocytosis of the bUSPIO-particles, since no receiving only bUSPIO suspension.
The administrated iron (500 ng) maximal concentration of 1,7 ng/µl
was small, as there have been studies with intracerebral administration of 25µg of iron (Muldoon et al., 2005). As compared to the reported detection limits of 10 µM, the concentration factor would be 1000-fold. Wh
fl (D
Baavi particles are likely to have esca
The dimishment of the MRI signal loss during time might be explained by CP epithelial cell turnover (Netsky and Shuangshoti, 1970; Chauhan and Lewis, 1979) or even cell proliferation as response to injury from the injection pressure (Li et al., 2002). After intracerebral administration the USPIO particles have been observed to accumulate in cervical lymph nodes via CSF efflux (Kida et al., 1993; Muldoon et al., 2004) which agreed with previous findings of baculovirus systemic escape to ectopic tissues (Lehtolainen et al., 2002) and data from later SPECT experiments (III / fig 3). These observations could explain the fate of free bUSPIO and bUSPIO-Baavi complexes.
The nonlinearity of the iron signal hinders the evaluati
Iron detection by Prussian blue staining
The relationship of the iron originated signal loss to the biodistribution ed by Prussian blue staining for iron. As the amount inobenzidine (DAB) signal enhancement for the detection (Moos a
thod resulted in staining of the cuboid epithelial cells of CP on the side of the injection, while the contralateral CP cells remained without any stain, similar to controls without any injected iron.
As the CSF flow originates from the lateral ventricles and passes ourth ventricle to the to superior sagittal sinus via the subarachnoid spac
liquid dynamics (Figure 13) are likely to prevent the diffusion of bUSPIO-virus pa ed ipsilateral ventricle to the contralateral ventricle. This would explain the lim of the coated virus.
5.2.5
of the baculovirus was
confirm of iron was low, we used
diam nd Mollgard, 1993). This
me
through the third and
f e (Knopf et al., 1995), the
rticles from the
inject ited diffusion
Figure 13. The flow of cerebrospinal fluid (CSF) in human brain, (a) coronal plane and (b) midsagittal plane.
The CSF is produced in ependymal cells of choroid plexus cells in lateral ventricles (1.) and it flows through the third ventricle (2.) to the fourth ventricle where it can continue to the spinal subarachnoidal space or (3.) continue through the lateral aperture of the fourth ventricle to the subarachnoidal space (4.) of the brain. The CSF in the subarachnoidal space moves to the superior sagittal sinus (5.) where it is reabsorbed via the arachnoidal villi into the venous system.
5.2.6 Viral transgene expression
In order to compare and examine the viral transgene expression, we performed β-galactosidase staining for the viral LacZ-transgene expression. As previously (Lehtolainen et al., 2002; Laitinen et al., 2005a), the β-galactosidase staining resulted in staining of cuboid epithelial cells of CP. Only a few blue cells were seen in the contralateral side, in agreement with the Prussian blue staining for iron. As compared to wild-type virus or uncoated Baavi, the results in ipsilateral side were similar or transduction efficiency was slightly increased, determined by the amount of blue cells. Since the USPIO coating increases the size of the virus, it is possible that larger sized complexes are more
easily endocytosed into the cells (Muro et al., 2004), as also suggested in a study which compared the uptake of different sized iron oxide particles (Raynal et al., 2004).
trast to the wild-type virus, some blue cells were detected sporadically in the
ination of the bUSPIO particles from CP, it would be interesting to determine if the similarity is caused by cellular regeneration.
inly in young children with 0.3 cases per a 1 million population. While surgical treatment is the primary option in the majority of the cases, the poor prognosis even with radio- and chemotherapy suggests that gene therapy could be used along with the traditional methods (Gupta, 2003; Wolff et al., 2002). There is also evidence that weight regulation via leptin interaction could be affected by numerous leptin receptors in CP (Strazielle and Ghersi-Egea, 2000), creating commercial potential for weight regulation by gene therapy.
In this light, the further research of the baculovirus mediated transduction of CP seems reasonable. The described MRI-based imaging has several benefits as compared to the traditional
In con
brain parenchyma with Baavi and USPIO-coated Baavi. It may be possible that the positive charge of the avidin on Baavi might enhance attachment and result in improved transduction of known permissible cell types, such as endothelial cells of microvessels (Lehtolainen et al., 2002), glial cells or astrocytes (Sarkis et al., 2000).
It is known that the expression of baculovirus transgene is reduced from 80 % at day 5 to almost zero at day 14 (Lehtolainen et al., 2002). Interestingly, the iron-related signal loss was also reduced to background in two weeks. Although short-time expression of baculovirus may be entirely separate from the elim
5.2.7 Choroid plexus as targets for gene therapy
This study showed that detection of viral particle biodistribution by MRI is possible. The data confirmed the co-localization of bUSPIO with Baavi and moreover indicated that Baavi could deliver cargo to CP cells. Magnetic nanoparticles have been utilized as drug carriers (Alexiou et al., 2006), with therapeutic results and decreased side effects (Alexiou et al., 2003). The combination of gene delivery and drug delivery together might be beneficial when considering the versatile role of CP in physiology.
CP are best known best for their role in producing CSF, but they also play a role in the immune system, maintaining the blood-brain barrier (BBB), detoxification, secretion of various molecules and neurogenesis (Emerich et al., 2005). Human CP cells produce 500 ml of CSF per day and are reported to be involved in various medical conditions, such as Alzheimer’s disease (Emerich et al., 2005), and in brain regeneration, containing neuronal precursor cells (Li et al., 2002). Therefore the possibility of affecting the central nervous system via the CSF by gene therapy would be intriguing. Interestingly, it has been suggested that 30% of CSF is produced at other sites than CP, such as the epithelial lining of the ventricles and the endothelium of the brain capillaries (Davson, 1972; Hammock and Milhorat, 1973). Together with the CP cells, all those sites have been found to be transduced with baculovirus (Lehtolainen et al., 2002; Kaikkonen et al., 2006).
Although this behaviour might be due to common extracellular properties deriving from embryonal properties (Sarnat, 1998), further research on this matter might also shed some light on to the baculoviral tropism and cell entry in the central nervous system (CNS).
According to literature, viruses such as Sendai-, mumps- and human T-cell leukaemia virus-1 and possibly HIV-1 have tropism for CP (Levine, 1987; Strazielle and Ghersi-Egea, 2000).
This might be explained by the role of CP in the neuroimmune system (Engelhardt et al., 2001), relaying information between the brain and the immune system (Lacroix et al., 1998b). Recently, CP transplants have been studied for neuroprotective potential and spinal neuron regeneration (Lacroix et al., 1998a). Ex vivo transduction of neuroprotective genes, such as VEGFs (Zachary, 2005) could be combined with such transplants to further enhance the neuroprotection.
CP tumours have a rare prevalence of 0.5 % of all brain tumours occurring ma
histochemical mechanisms. Especially when including a MRI-visible transgene system with ifferent relaxivity (Table 5) to the virus, both particle biodistribution and transgene imaging could
erfusion and metabolic and ctions (Yang and Atalar, 200 , this MRI-based baculovirus imaging system to further characterize novel t or CP related conditions.
d
be performed with the same device, although iron effects to detection of other contrast agents as well. Additionally, MRI is also capable of imaging physiological changes in organ p
mechanical fun
could be used 6). In time
reatments f
5.3 Article III
As discussed in the previous chapter, the transduction of choroid plexus in brain has so far showed promising results and set goals for further studies. As intracerebroventricular (ICV) administration is mostly suited to the treatment of brain disorders, there is still little knowledge of the baculoviral systemic kinetics in the body after ICV injection and traffic routes after other administration routes.
Since the baculovirus is hindered by the blood complement (Hofmann and Strauss, 1998), biodistribution studies based on the transduction pattern of the virus would not provide accurate
formation about the viral kinetics after administration, but tracking viral particles could provide dditional information about baculovirus systemic kinetics.
with biotinylated USPIO could create ution, the poor temporal resolution favour the
idney and a minor
99m
resulted b
virus with the labelled chelate did not show any activity. Despite the SPECT revealed activity in the spleen, no beta-galactosidase expression was detected with Baavi or wild-type virus.
in a
While the MRI- imaging of Baavi coated accurate information about brain particle biodistrib
use of other imaging method when acquiring data of rapid systemic changes in baculovirus administration kinetics. As discussed in 2.4.1.2, SPECT imaging has these advantages over MRI imaging, also enabling quantitation of the signal.
In this study, we compared different routes of administration of avidin-displaying baculovirus, Baavi, coated with polylysine-serine-DTPA chelate (III/ fig 1) with 99mTechnetium. By using a combined microSPECT/CT device the particle biodistribution could be monitored real-time in the SPECT with the ability to include anatomical references by CT imaging.
5.3.1 Intravenous administration
Systemic injection is the preferred method to use, when acquiring a wide access to organs by gene delivery vectors. The complement inhibition hinders the systemic use of baculovirus; however the inhibition is not always complete. Previously it has been shown that systemic administration of non-modified baculovirus through a tail vein injection in BALB/c mice resulted in GFP expression in liver, spleen, lung, heart, kidney and brains (Kim et al., 2006). Similarly, after systemic injection of A/J mice a significant lusiferase expression was seen in the spleen, liver and k
expression in the lungs (Kircheis et al., 2001). However, no GFP or luciferase expression was seen after intravenous administration of VSV-G displaying baculovirus into BALB/c mice (Tani et al., 2003b). While the results may be due to pseudotyping of virus, a variation in dose, rodent strain, individual properties of experimental animals or virus preparation between laboratories, there might also be other factors influencing to the expression pattern.
The SPECT planar imaging of Baavi coated with technetium-labelled chelate showed increasing activity after systemic injection via v. femoralis, (IF), (III/ fig 2) in the lungs, liver, spleen and kidneys while the imaging of the radiolabelled chelate or Tc alone (data not shown) in rightening of kidneys, ureters and bladder as the urinary export of water-soluble substances progressed. It has been shown that the reticulo-endothelial system of the liver plays an important role in the active disposal of administered viral particles (Tao et al., 2001), together with the spleen and kidneys, being therefore expected to gather virions. The signal seen in the lungs was decreased in time, likely representing viruses circulating in blood (Schellingerhout and Bogdanov, Jr., 2002).
Baavi resulted in moderate beta-galactosidase expression in kidneys and liver (III, table 1), while IF injections of the wild-type
5.3.2 traperitoneal and intramuscular administration
traperitoneal administration (IP) has been used for the treatment of ovarian cancer, in an attempt proved access to the peritoneal cavity has a direct access to the thoracic and
transduce various mammalian kidney cell lines in vitro (Liang et al., 2004;
Condreay et al., 1999). The reports of kidney transduction after systemic tail-vein administration do
l, although not clinically preferred method. Previously it has been reported that the intracranial administration of baculoviral vector has resulted in GFP expression in mouse and rat brain striatum, corpus callosum and ependymal layer neuronal cells
GFP expression in the mouse brain (Tani et al., 2003b), In
In
to overcome the limitations of intratumoral injections for im (Evans and Keith, 2004). However, the peritoneal cavity
pleural cavities via the lymphatic channels (bu-Hijleh et al., 1995a). The peritoneal fluid enters from the peritoneum to the lymphatic lacunae and continues onwards via the parasternal lymphatics to the superior mediastinal nodes (bu-Hijleh et al., 1995b; bu-Hijleh et al., 1995a). This access enables both ovarian tumour metastases and injected virus spread systemically through the lymphatic system, the latter being important for viral safety.
IP injection of the labelled Baavi resulted in significantly increased radioactivity in spleen and kidney (III/ table 1). Beta-galactosidase stainings of the spleen resulted in moderate expression of the LacZ transgene and in the kidney resulted in strong and extensive beta-galactosidase expression (III/ Figure 4). According to a study, baculoviruses have previously been shown to efficiently
not show extensive transduction in one study (Kim et al., 2006), while the other shows kidneys quite positive (Kircheis et al., 2001). It might be possible that avidin display has enhanced the baculovirus transduction mechanism as compared to the wild-type virus, as suggested by our previous study (I), to result in the successful transduction of the kidney after both IP and IF administrations.
After IP administration some beta-galactosidase expression could also be seen in the lungs and the brain and radioactivity was occasionally detected in mediastinal nodes during the SPECT imaging (data not shown). Interestingly, the brain was also seen unexpectedly positive after systemic administration in a previous study, increasing with the PEGylation of the baculovirus (Kim et al., 2006). This indicates that under suitable conditions the baculovirus is possibly able to transport across the BBB and transduce brain parenchymal cells with better capability than previously thought. Our results suggest that after IP injection the baculovirus enters the lymphatic circulation from the peritoneal lymphatic drainage. Similarly after intramuscular injection with Baavi, we could detect the femoral lymphatic nodes being imaged, indicating viral traffic via the
After IP administration some beta-galactosidase expression could also be seen in the lungs and the brain and radioactivity was occasionally detected in mediastinal nodes during the SPECT imaging (data not shown). Interestingly, the brain was also seen unexpectedly positive after systemic administration in a previous study, increasing with the PEGylation of the baculovirus (Kim et al., 2006). This indicates that under suitable conditions the baculovirus is possibly able to transport across the BBB and transduce brain parenchymal cells with better capability than previously thought. Our results suggest that after IP injection the baculovirus enters the lymphatic circulation from the peritoneal lymphatic drainage. Similarly after intramuscular injection with Baavi, we could detect the femoral lymphatic nodes being imaged, indicating viral traffic via the