Formation of Microbubbles for Targeted Ultrasound Contrast Imaging

Subscriber access provided by TULANE UNIVERSITY

Interface Components: Nanoparticles, Colloids, Emulsions, Surfactants, Proteins, Polymers

Formation of Microbubbles for Targeted Ultrasound Contrast Imaging: Practical Translation Considerations. Sunil Unnikrishnan, Zhongmin Du, Galina Diakova, and Alexander L Klibanov Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.8b03551 • Publication Date (Web): 03 Dec 2018 Downloaded from http://pubs.acs.org on December 7, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 47 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Formation of Microbubbles for Targeted Ultrasound Contrast Imaging: Practical Translation Considerations. Sunil Unnikrishnan2, Zhongmin Du1, Galina B Diakova1, Alexander L Klibanov1,2*.

1Cardiovascular

Division, Department of Medicine, Robert M. Berne

Cardiovascular Research Center, University of Virginia School of Medicine 2Department

of Biomedical Engineering, University of Virginia, Charlottesville VA 22908.

*Corresponding Author: Alexander L Klibanov, UVA Cardiovascular Division, CVRC, MR4 RM3147, 409 Lane Rd., Charlottesville VA 22908-1394. Email: [email protected]; [email protected].

ACS Paragon Plus Environment

1

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 47

Key Words: Microbubbles; Contrast Ultrasound; Molecular Imaging; Ultrasound Imaging; Targeted Microbubbles; Targeting; VCAM-1; V3; Lipid Monolayer; Decafluorobutane.

Abstract.

For

preparation

of

ligand-decorated

microbubbles

for

targeted

ultrasound

contrast imaging it is important to maximize the amount of ligand associated with the bubble shell. We describe optimization of the use of a biocompatible co-surfactant

in

the

microbubble

formulation

media

to

maximize

the

incorporation of targeting ligand-lipid conjugate into the microbubble shell, and thus reduce the fraction of ligand not associated with microbubbles, following amalgamation preparation. The influence of the concentration of a helper co-


2

Page 3 of 47 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

surfactant propyleneglycol (PG) on the efficacy of microbubble preparation by amalgamation and on the degree of association of fluorescent PEG-lipid with the microbubble shell was tested. Three sets of targeted bubbles were then prepared: with VCAM-1-targeting peptide VHPKQHRGGSK(FITC)GC-PEG-DSPE, cyclic RGDfK-PEG-DSPE, selective for V3, and control cRADfK-PEG-DSPE, without such affinity. Microbubbles were prepared by 45s amalgamation, with DSPC and PEG stearate as the main components of the shell, with 15% PG in aqueous saline. In vitro microbubble targeting was assessed with a parallel plate flow chamber with a recombinant receptor coated surface. In vivo targeting was assessed in MC-38 tumor-bearing mice (subcutaneous tumor in hind leg), 10 min after intravenous bolus of microbubble contrast agent (20 million particles per injection). Ultrasound imaging of the tumor and control nontarget muscle tissue in a contralateral leg was performed with a clinical scanner. Amalgamation technique with PG co-surfactant produced microbubbles at concentrations exceeding 2.109 particles per ml, and ~50-60% or more of the added

fluorescein-PEG-DSPE

or

VCAM-1-targeted


fluorescent

peptide

was

3

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 47

associated with microbubbles, about 2 times higher than in the absence of PG. After intravenous injection, peptide-targeted bubbles selectively accumulated in the tumor vasculature, with negligible accumulation in non-tumor contralateral leg muscle, or with control non-targeted microbubbles (assessed by contrast ultrasound

imaging).

For

comparison,

administration

of

RGD-decorated

microbubbles prepared by traditional sonication, and purified from free peptidePEG-lipid by repeated centrifugation, resulted in the same accumulation pattern as for translatable amalgamated microbubbles. Following amalgamation in the presence of PG, efficient transfer of ligand-PEG-lipid to microbubble shell was achieved and quantified. Purification of microbubbles from free peptide-PEG-lipid was not necessary, as proven by in vitro and in vivo targeting studies, so PG co-surfactant amalgamation technique generated peptide-targeted microbubbles amenable for bedside preparation and clinical translation. Pathway to clinical translation is simplified by the fact that most of the materials used in this study are either on FDA GRAS list, or can be procured as pharmaceutical grade substances.


4

Page 5 of 47 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Introduction. Molecular ultrasound imaging with targeted microbubble contrast agents is gradually making transition from initial preclinical testing

1

to Phase 1 and

Phase 2 clinical trials 2. In clinic these contrast agents may eventually provide precise location of lesions to guide biopsy or ablation ischemic memory condition

4,

3

or rapidly diagnose

within minutes of contrast administration, at

patient bedside, and without the need for expensive and time-consuming PET, SPECT or MRI clinical examinations. The basis for such contrast ultrasound studies is a gas-filled microbubble, a particle that is typically filled with a perfluorinated gas, and a thin shell, which is most often a lipid monolayer. As microbubble particles are intravascular blood pool contrast agents, their size should not exceed the diameter of blood capillaries. Yet acoustic backscatter signal is much higher for larger bubbles (Rayleigh scattering is proportional to the sixths power of bubble diameter). Therefore, most of the studies are performed with particles that have mean diameter ~1.5-4 um, which is sufficient


5

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 47

to visualize single microbubbles in real time, several cm deep in the tissues of interest 5. Such microbubble particles have been traditionally prepared by sonication where perfluorinated gas is sparged through the aqueous micellar medium, and a probe-type sonicator is used to break gas phase into fragments that form microbubbles. Perfluorinated gases are preferred due to very low solubility in water, which improves microbubble stability. Such microbubble preparation, which entrap e.g., perfluorobutane gas, with proper shell selection, e.g., DSPC, if sealed under perfluorobutane atmosphere, can be stored for many months, in the aqueous medium; they can carry biotin residues attached to the shell via a PEG spacer arm 6. A significant problem is the difficulty with coupling yield efficacy for targeting ligand attachment to microbubbles. For example, during the early preclinical studies, covalent attachment of anti-ICAM-1 to the microbubble shell via carbodiimide coupling chemistry

1b

was extremely inefficient; well over 90% of

the added (very expensive) antibody was lost, which is not appropriate for


6

Page 7 of 47 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

practical translation. Therefore, biotinylated microbubbles were used widely since the initial preclinical in vivo microbubble targeting study

1c,

where biotinylated

targeting ligands (antibodies against P-selectin) were attached via a streptavidin bridge with an excellent yield. This approach became universally popular, with many biotinylated antibodies glycosulfopeptides

8

7

or other biotinylated targeting ligands, such as

or carbohydrate-based ligands

4, 9,

attached to biotinylated

microbubbles. However, clinical translation with the use of avidin or streptavidin - foreign proteins immunogenic to humans - is not possible. Thus, a more robust and practical techniques for microbubble preparation and targeting ligand attachment are desired. One approach would be to prepare a covalent conjugate of a targeting ligand first, via e.g., a poly(ethylene glycol) (PEG) spacer to a lipid anchor, e.g., distearoyl phosphatidylethanolamine (DSPE), in optimal conditions, with high yield. Then this conjugate is admixed with other lipids normally used for microbubble preparation (such as DSPC and PEG stearate), and microbubbles are made by the sonication technique

6,

where

perfluorocarbon gas is rapidly dispersed in the aqueous medium that contains


7

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 47

shell-forming lipids in micellar form, and microbubbles are generated at high particle concentration. Unfortunately, following this sonication procedure, most of shell materials, including ligand-PEG-lipid, do not associate with the surface of the resulting microbubbles: they are retained in the micellar form in the aqueous medium. This inefficiency mandates centrifugal separation of free lipids from the bubbles, and makes clinical translation non-practical. An alternative approach - preparation of microbubbles by vibration/amalgamation, originally proposed by Unger, Fritz, Matsunaga et al.,

10,

when a sealed vial with

aqueous lipid medium and perfluorocarbon atmosphere is agitated in a dental shaker, yields similar results in many experimental conditions. The use of additional helper molecules, which possess a moderate degree of surfactant activity and assist with microbubble generation is deemed necessary

10-11.

In

this study, we investigate an initial optimization of targeted microbubble formulations, using a biocompatible helper molecule, propylene glycol (PG). The goal is to achieve and confirm high degree of transfer of the present micellar lipid material onto the microbubble shell. At the same time we need to assure


8

Page 9 of 47 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

quick and efficient preparation of targeted microbubbles, at concentrations sufficient for practical use. The aim is to avoid the undesirable purification procedures

that

remove

formulation,

streamlining

free

unincorporated

practical

application.

targeting Clinical

ligand

translation

from of

the such

formulations may therefore become feasible.

Experimental Section. Materials. Distearoyl phosphatidylcholine (DSPC) and DSPE were obtained from Avanti Polar Lipids, Alabaster, AL. PEG6000 monostearate was from Stepan Kessco (Northfield IL). NHS-PEG3400-DSPE and fluorescein-PEG5000NHS were from Shearwater Polymers (Birmingham, AL). Maleimide-PEG3400DSPE

was

Chemicals

from

Laysan

(Preston,

Bio

UK).

(Arab,

AL).

Peptides

Perfluorobutane were

either

was

from

F2

custom-ordered

(VHPKQHRGGSK(FITC)GC, Lifetein, Somerset, NJ), or obtained from Bachem, Torrance CA (cRGDfK and cRADfK). Recombinant V3 was from R&D Systems (Minneapolis, MN). Inorganic salts, buffers and organic solvents were from


9

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 47

Sigma-Aldrich (St. Louis, MO) or Fisher (Waltham MA). Dulbecco’s PBS (CaMg-free) and DMEM tissue culture media were from GIBCO (Grand Island, NY). Normal saline (0.9% NaCl) was from Baxter (Deerfield IL).

Methods. Fluorescein-PEG-DSPE

preparation.

Fluorescein-labeled

PEG-DSPE

was

prepared by adding 1.2x molar excess of fluorescein-PEG5000-NHS to DSPE in DIPEA-containing chloroform. After completion of the reaction, organic solvents were removed under vacuum, material redissolved in saline and dialyzed against saline and then against water, in large-pore 100 KDa MWCO dialysis bags (Spectrum Laboratories, Rancho Dominguez, CA) to remove excess of PEG from PEG-DSPE micelles. After completion of dialysis, micellar fluoresceinPEG-DSPE was lyophilized and stored frozen. Peptide-PEG-DSPE conjugate preparation.

12.

Micellar aqueous solution of

maleimide-PEG-DSPE in degassed PBS buffer containing 0.1mM EDTA was added to solid VHPKQHRGGSK(FITC)GC thiol under argon gas atmosphere,


10

Page 11 of 47 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

with 1.1x excess of peptide, and incubated overnight in the dark. Excess of unreacted peptide was removed by sequential dialysis against normal saline and water, and final product lyophilized. Cyclic peptides, cRGDfK (with affinity to v3)

13,

or a non-binding control

cRADfK, possessing a single primary aminogroup for active ester covalent coupling,

were

mixed

with

equimolar

amount

of

NHS-PEG3400-DSPE

in

DMSO/chloroform (at 5mg/ml), and 1.2x molar excess of DIPEA. The mixture was incubated until completion of the reaction as monitored by ninhydrin spray on TLC plates. Volatile organic solvents were removed by prolonged highvacuum lyophilization, and material used without further purification. Preparation of microbubbles by sonication. A micellar dispersion in normal saline was prepared by probe-type sonication (XL2020, Misonix, Farmingdale, NY) equipped with half-inch extended probe, “3” power setting) until reduction of dispersion turbidity to almost complete clarity. Micelles contained DSPC (2 mg/ml), PEG stearate (2 mg/ml) and ligand-PEG-DSPE (0.1 mg/ml). To generate

microbubbles,

a

flow

of

perfluorobutane


gas

through

the

lipid

11

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 47

dispersion was initiated via a teflon capillary tubing, to completely replace air in the beaker headspace, and then was subjected to sonication with XL2020 unit, (30 sec at maximum power, to generate microbubbles). Resulting microbubble dispersion was rapidly cooled in ice water, dispensed in glass vials and sealed under perfluorocarbon atmosphere for extended refrigerated storage at 4oC. On the day of the study, microbubble samples were placed in syringes and subjected

to

repeated

(4x)

centrifugation

in

de-aerated

decafluorobutane-

saturated saline to remove unincorporated lipids and ligand (10 min, 1000 rpm, HN-SII centrifuge with a bucket rotor, ~100 x g). Next, flotation for 10 min was used

to

remove

large

microbubbles

from

the

preparation.

Microbubble

concentration and size were determined with the use of Coulter Multisizer 3 counter (Beckman Coulter, Hialeah, FL) in 1-30 um range. Microbubble preparation by amalgamation: fluorescein-PEG-DSPE attachment. A mixture of DSPC and PEG stearate was dispersed in PBS by sonication, and Flu-PEG-DSPE solution in PG was added. The mass ratio between DSPC and fluorescent lipid was maintained at 20:1 (molar ratio ~154:1). Into PBS-


12

Page 13 of 47 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

lipid-containing vials, PG, pre-heated in a water bath to improve handling, was added,

ranging

between

0.25%

and

30%

final

v/v

concentration,

and

supplemented with PBS to equalize the final volume in all samples. In one of the sample sets, DSPC and PEG stearate concentrations were 0.25 mg/ml; in another set, they were at 1 mg/ml. The vials were put in a hot water bath to assist distribution of lipids into PBS-PG aqueous medium. Vial headspace was filled with perfluorobutane gas, vials were stoppered, sealed and crimped. Vials were allowed to reach room temperature prior to amalgamation, which was performed at 4300 rpm for 45 s with a Vialmix apparatus (Lantheus, North Billerica, MA). Coulter counter was used for microbubble size and concentration determination (see above). Microbubble

preparation

by

amalgamation:

VCAM-1-targeted

peptide-PEG-

DSPE attachment. Solutions of DSPC and PEG stearate (1:1 mass ratio) in PG as well as solution of VHPKQHRGGSK(FITC)GC-PEG-DSPE in PG were incubated in hot water bath and mixed to maintain 5:1 DSPC: peptide-PEGDSPE mass ratio (~8:1 molar ratio). PG and saline were added, to maintain


13

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 47

final PG concentration at 15%. DSPC concentration was varied between 0.25mg/ml and 2 mg/ml. Microbubble preparation by amalgamation: αVβ3-targeted peptide attachment. Solutions of DSPC and PEG stearate (1:1 mass ratio) in PG were mixed with cRGDfK-PEG-DSPE or control cRADfK-PEG-DSPE (DSPC:peptide-PEG-DSPE mass ratio was kept constant at 20:1, i.e., molar ratio was ~120:1), and added to normal saline in 2 ml vials in a hot water bath, to achieve the final concentration of 1 mg/ml DSPC. PG was added to achieve 15% final concentration. Vials were filled with perfluorobutane gas headspace, sealed, stoppered and crimped for refrigerated storage. Determination of the fraction of label-PEG-lipid associated with microbubbles. Following

amalgamation,

the

sealed

vial

containing

microbubbles

with

fluorescent label (fluorescein-PEG-DSPE or FITC-carrying VCAM-1-targeted

14

peptide-PEG-DSPE) was turned upside down, placed in a centrifuge tube and spun for 10 min in a bucket rotor of HN-SII centrifuge (IEC, Thermo Scientific, Waltham, MA) at 1000 rpm (~100 x g). After completion of centrifugation, as


14

Page 15 of 47 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

microbubbles collected at the top of the saline medium, a sample of infranatant liquid was collected from the inverted vial, close to the septum, using a syringe with a thin short needle. Sample aliquots were placed in 96-well plates (Costar black, flat bottom, Corning, Kennebunk ME), diluted in PBS and subjected to fluorescence spectroscopy (485 nm excitation, 535 nm emission, Spectramax Gemini

XS

microplate

reader,

Molecular

Devices,

Sunnyvale,

CA),

and

compared with the signal from the samples prior to amalgamation to determine the fraction of fluorescent material extracted from the bulk liquid into the upper (microbubble) phase.

In vitro microbubble targeting. control

cRADfK-PEG-DSPE

cRGDfK-PEG-DSPE-carrying microbubbles, or

microbubbles

were

tested

for

targeting

to

recombinant murine αVβ3 in a parallel plate flow chamber system, essentially as described

earlier

15

Briefly,

recombinant

target

protein

was

dissolved

in

Dulbecco’s PBS at 4 ug/ml, 0.2 ml drop placed in the center of a standard 10x35mm polystyrene Petri dish (Corning Inc., Corning, NY), the drop covered with a polymer coverslip for even distribution, dishes sealed in a moist chamber


15

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 47

and incubated overnight to achieve target protein adhesion to the surface. Residual nonspecific binding sites on the dish surface were then blocked by incubation with BSA solution (0.2% in PBS). As another control, BSA-only plates were used. A 35mm circular plug parallel plate flow chamber (Glycotech, Gaithersburg, MD) was used. Microbubble dispersion (5.106/ml) was flown through the chamber with the aid of a syringe pump (Harvard Apparatus, Holliston MA) to achieve 1 dyn/cm2 wall shear stress, under video microscopy observation (Leica Laborlux 11 compound microscope equipped with 40x objective and an NTSC video camera, connected to a PC via a digitizer). Microbubble accumulation in the field of view at the target or control surfaces was quantified and compared after 4 min of microbubble flow through the chamber; statistical significance was assessed with Student’s t-test.

In vivo microbubble targeting. In vivo microbubble targeting experiments were performed in accordance with protocol approved by the University of Virginia’s Institutional

Animal

Care

and

Use

Committee.

MC38

murine

colon

adenocarcinoma cells, generously provided by Dr. J. Schlom (NIH, Bethesda,


16

Page 17 of 47 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

MD) were grown in high-glucose DMEM medium (GIBCO) supplemented with pyruvate, 10% heat-inactivated FCS, penicillin and streptomycin. Cells were removed from the flasks by gentle pipetting, washed in PBS, and injected subcutaneously (500,000 cells per animal) in the hind leg of C57BL/6 mice (NCI, Bethesda, MD). Tumors were grown until they reached ~1 cm in size. For all manipulations mice were under isoflurane anesthesia (1.5-2% in air). The animals were placed on a warm pad. Hair was then removed from the skin above the tumor and at the control contralateral leg area. Microbubbles in normal saline (20 million) were injected intravenously as a bolus; contrast ultrasound imaging was performed with the use of 15L8 transducer of an Acuson Sequoia c512 scanner operated at low transmit power (MI 0.2, 7 MHz) in intermittent mode (1 Hz), to minimize chances for microbubble destruction by ultrasound. The transducer was placed to observe the acoustic backscatter signal from the tumor tissue and control contralateral leg muscle at the same time. Ultrasound imaging video stream was recorded and frames digitized with ImageJ software (NIH, Bethesda, MD). Targeted microbubble adhesion was


17

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 47

assessed 10 min after the microbubble contrast bolus injection, as a fraction of the echo contrast signal (mean pixel intensity at the region of interest) recorded at peak (30 seconds). Statistical significance was assessed with Student’s t-test.

Results. Assessment of microbubble preparation efficiency by sonication. Traditional sonication preparation of microbubbles is based on the original albumin solution sonication technique

16,

where air or another gas

17

is dispersed in hot albumin

solution in aqueous medium by probe-type sonication. This technique is in use to generate lipid and surfactant bubbles for at least two decades

18 1b.

In the

study reported here we assessed the lipid concentration dependence on the efficacy of this sonication method to generate microbubbles using lipid (DSPC) or a lipid-cosurfactant mix (DSPC-PEG stearate). In the absence of PEG stearate,

microbubble

formulation

was

highly

inefficient

and

the

resulting

microbubble concentration was very low. It slightly exceeded 20 million particles per ml only at the highest concentrations of DSPC (Figure 1, open circles).


18

Page 19 of 47

5E+09

particle concentration

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

5E+08

5E+07

5E+06 0

0.5

1

1.5

2

DSPC, mg/ml Figure 1. Generation of decafluorobutane microbubbles by probe-type sonication of the aqueous lipid dispersion. The influence of DSPC lipid concentration on the resulting microbubble concentration, in the absence (open circles) and in the presence (closed diamonds) of PEG stearate cosurfactant (billions per ml).

This

product

concentration

would

not

be

useful

for

practical

application

purposes: clinical grade bubbles had always been reported as having above


19

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 47

500 million particles per ml. The fraction of lipid transferred to the microbubble shell was also extremely low. With the addition of PEG stearate at 1:1 mass ratio to DSPC (molar ratio ~1:8), microbubble formulation became practically feasible, especially at higher lipid concentrations, (Figure 1, closed diamonds, note logarithmic scale), exceeding 3.5.109 particles per ml, almost 200-fold higher than without PEG stearate. Without any additional purification, 98% of the particles were under 10 um in diameter. However, even in the presence of PEG stearate, less than a quarter of added lipid was transferred from the aqueous phase to the microbubble shell. Therefore, further optimization was required. It was accomplished by switching to the amalgamation preparation technique.

Optimization of the microbubble amalgamation medium. As the first step in the microbubble amalgamation optimization, we investigated the influence of PG concentration on the efficacy and size of microbubble formation and transfer of fluorescein-PEG-DSPE to the bubble shell, at two concentrations of the main


20

Page 21 of 47 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

shell lipid, DSPC, 0.25 mg/ml and 1 mg/ml. At 0.25 mg/ml and low PG concentrations (2% or less), very limited fluorescence material transfer to the bubbles occurred, and the resulting bubble concentration following amalgamation at that condition was insufficient for practical use, under 5.108 particles per ml. At higher PG concentrations, transfer of fluorescent peptide mimic reached and exceeded 40% (Figure 2, diamonds), and bubble concentration exceeded 1.5.109/ml. The fraction of undesirable large bubbles (over 10 um in diameter) was less than 1% of total; at 15% PG it was ~0.12%. At 1 mg/ml DSPC concentration (Figure 2, triangles), the preparation was even more efficient: dye transfer to bubbles was maintained between 60 and 70% for 10-25% PG concentration range. Bubble concentrations in those conditions exceeded 2.109 particles/ml, and most of the generated bubbles were in the desired size range, with only 0.62% above 10 um in diameter for 15%PG.


21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

% inclusion of flu-PEG-DSPE in MB

Langmuir

Page 22 of 47

70 60 50 40 30 0.25 mg/ml DSPC

20

1 mg/ml DSPC

10 0 0

Figure

2.

Influence associated

5 Generation

of

PG with

10

20

PG concentration, % vol of

microbubbles

concentration the

15

on

microbubble

the

by

amalgamation

fraction

shell,

of

25

30

(Vialmix,

45

s).

fluorescein-PEG-DSPE

determined

with

fluorescence

spectroscopy. DSPC concentration in the medium was 0.25 mg/ml (diamonds) or 1 mg/ml (triangles). Mass ratio of DSPC: PEG stearate:fluorescein-PEGDSPE, 20:20:1 ( molar ratio ~154:19:1).


22

Page 23 of 47

Based on this initial optimization, we investigated the influence of lipid concentration on the efficacy of transfer of the actual VCAM-1-targeting peptidePEG-DSPE conjugate from the amalgamation medium to microbubble shell, and the importance of PG presence. It was found that while increasing the total lipid concentration improved ligand-lipid conjugate transfer to the microbubble shell, the presence of 15% PG enhanced transfer efficacy even further: it reached ~60-70% of the added peptide (Figure 3).

% inclusion of VCAM-1-peptide

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

70 60 50 40 30

no PG

20

15 vol.% PG

10 0 0

0.5

1

1.5

2

mg/ml DSPC ACS Paragon Plus Environment

23

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure

3.

Generation

of

microbubbles

Page 24 of 47

by

amalgamation

(Vialmix,

45

s).

Influence of lipid concentration on the fraction of VHPKQHRGGSK(FITC)GCPEG-DSPE associated with the microbubble shell, determined with fluorescence spectroscopy. PG concentration, 15% (diamonds), or none (circles). Mass ratio of DSPC:PEG stearate:peptide-PEG-DSPE 5:5:1 (molar ratio ~36:4.5:1).

For further testing of peptide-mediated microbubble targeting in vivo we selected

samples

amalgamation

in

based

on

these

15%

PG

conditions,

and

1

particle

mg/ml

DSPC.

Following

concentrations

(>>1.109

microbubbles/ml) were sufficient for testing of microbubble targeting in vivo after a minor adjustment of size distribution (removal of the subset of potentially interfering larger bubbles) by a 10 min flotation incubation of inverted vials at normal gravity, followed by collection of the desired fraction of size-refined microbubbles (typically, ~60% of total volume) from the bottom of the vial through the septum with a short needle into a syringe for subsequent testing of targeting in vitro and in vivo.


24

Page 25 of 47

In vitro peptide-mediated microbubble targeting. An in vitro flow chamber experiment

was

used

to

assess

targeted

adhesion

of

amalgamated

microbubbles. Peptide-decorated bubbles were used without removal of the residual unincorporated peptide ligands. Figure 4 shows that while RGDdecorated microbubbles efficiently adhered to the recombinant v3-coated surface, attachment of control non-targeted cRADfK (scrambled) microbubbles was minimal (p

Formation of Microbubbles for Targeted Ultrasound Contrast Imaging

Recommend Documents