"Fantastic Voyages," Feature Article, October 2006

Focus on Bioengineering

fantastic voyages

Nanodevices in development today promise to give medicine capabilities that were once purely in the realm of fiction.

by Paolo Decuzzi and Mauro Ferrari

advances in nanotechnology and biotechnology in the last few years have led to the development of a large variety of applications in diverse fields, from analytical chemistry to advanced materials.

Nano-biotechnology is expected to have still greater societal impact in the future. The main challenges in the field are to make significant advances in the detection, treatment, and imaging of diseases, such as cancer and cardiovascular diseases, which constitute by far the most common causes of suffering and death in industrialized countries.

In the early detection of diseases, the challenge is to design and develop devices with the ability to detect very small amounts—as small as a few pico-grams—of specific molecules, or biomarkers, in a background of thousands of other molecules dispersed in the blood. Biomarkers are released by abnormal cells, as tumor or infarcted myocardial cells, and as such they constitute the biological footprints of a disease.

Nanotechnology makes it possible to envision new devices that can deliver a 100-fold and even larger increase in sensitivity over current diagnostic techniques, resulting in the detection of disease at its earliest stage, favoring the diagnosis and subsequent treatment.

Our research group, a cooperative effort of the University of Magna Graecia in Italy and the University of Texas in Houston, in collaboration with other Italian universities, is currently focused on the development of silicon-based nanoporous particles. These are micrometer and submicrometer hemispherical particles fabricated following standard photolithographic procedures and made porous by anodization in an electrolytic solution. The pore size can be varied between 20 and 100 nm. The surface chemistry of the particles can also be controlled changing functional groups and electrostatic charge.

An injection of tiny drug-carrying devices able to seek out tumors and treat only diseased cells would avoid many of the side effects of chemotherapy.

These particles can be directly injected into the blood flow, through intravenous infusion, and selectively uptake the desired biomarkers over the "noise" of the most abundant blood proteins. Their selectivity can be significantly increased by tuning the pores' surface density and characteristic size, as well as the surface chemical properties of the particles relative to the biomarkers of interest. The particles are later extracted for analysis.

These harvesting nanoparticles have the advantage over other approaches of operating in vivo.

Evidently, these particles have to navigate within the whole circulatory system without adhering to a blood vessel wall or leaving the blood pool by extravasation. The vasculature appears as a complicated network of channels with different sizes and lengths, arranged in series and in parallel.

Navigating the System

How do particles choose their routes within such a tremendously intricate system? In the absence of a fluid flow, the motion of a particle is solely governed by Brownian diffusion. Within a capillary, the motion of the particles is also governed by the hydrodynamic forces exerted by the flowing blood on the particle surface. These forces depend on the size of the particle relative to that of the capillary, on the shape of the particle, on the hemodynamic conditions, and on the possible interactions of the particle with the corpuscular components of blood—white blood cells with a characteristic size of 15 micrometers, red blood cells averaging 10 µm, and platelets of approximately 5 µm—which tend to concentrate in the central zone of a capillary.

Making use of a theoretical analysis based on the dispersion of passive tracers in a confined laminar flow, we have shown that for a given capillary radius and hemodynamic condition there exists a critical particle size for which the longitudinal diffusion of the particle along a capillary is at a minimum. These critical-size particles would require the longest time to move along the capillary compared to larger or smaller particles. The size is affected by the permeability of the blood capillaries, being larger for the leaky capillaries, such as those encountered in tumor microvasculature.

As with heat or electric current, circulating particles will follow the less-resistive path. Therefore, particles with a critical radius for normal (nontumoral) blood vessels, that is, with a characteristic size of about 100 to 200 nm, will pass through tumor capillaries more readily and in greater concentrations than through healthy channels.

A real multiscale model for analyzing the transport of particles within blood capillaries is lacking: a model comprising information on the complex architecture of millimeter-long vessels, within which platelets and blood cells move with a velocity of 1 to 10 millimeters a second, and where submicrometer particles may interact with the blood vessel walls and thousands of soluble nanometer molecules. The coupling of standard macroscale continuum models of incompressible fluid flow with mesoscale discrete particle models can provide a clue to future developments in this field.

In the therapy of diseases, the challenge is to design and develop devices with the ability to deliver multiple therapeutic agents (drug molecules) with high selectivity and a precise release protocol. Selectivity in delivery implies that the drug molecules kill solely the abnormal cells and limit their interaction with the normal cells to reduce side effects, such as hair loss or nausea common to cancer chemotherapy.

A precise release protocol implies that the amount of administered drug molecules is kept within the optimal therapeutic levels. In current chemotherapy, where the therapeutic agents are directly injected into the bloodstream, only between 1 and 10 parts per 100,000 of drug molecules reach the tumor microenvironment—a negligible percentage, an engineer would say sadly.

Two main categories of drug-delivery devices have been identified: implantable devices and circulating devices based on the use of nanoparticles. Regarding the circulating devices, two delivery strategies can be followed: vascular targeting, where the nanoparticle is designed to adhere firmly to the cells of the tumor blood vessels; and tumor microenvironment targeting, where the nanoparticle is designed to recognize and adhere to the tumor cells in the tissue surrounding blood vessels.

Our group is currently developing a hybrid strategy: the double-stage particle. This is a particle with a characteristic volume of about 1 micrometer whose porous structure retains thousands of smaller particles with a characteristic size between 20 and 50 nm, which are the real carriers of the therapeutic agents. If even one of these particles adheres at the walls of a tumor capillary, 100,000 drug molecules can be released just within the tumor microenvironment to deliver a high and concentrated drug dose.

The adhesion of the particle to the cell is mediated by specific and non-specific interactions. At the cell-particle interface, stable non-covalent bonds are formed between molecules distributed over the particle surface (ligands) and countermolecules expressed over the cell surface (receptors).

From the analysis of the cell-particle adhesion under flow, we have shown that the size and the shape of the particles, as well as their surface chemistry, play a major role in the recognition and final firm adhesion to the biological target. Oblate spheroidal particles, which offer the largest surface area to the cells of the blood vessels and the smallest surface to the flowing blood can adhere more avidly to the vascular target and from there release their payloads. Also, such biomechanical properties of the ligand-receptor pair as the rupture load and the extent of stretch required to reach this load can be tailored to control the strength of adhesion. Chemistry and computations at the molecular scale can help design molecules with a precise mechanical strength.

Imaging, Perhaps Cell by Cell

In bioimaging, the challenge is to design and develop in-vivo devices with the ability to improve the spatial resolution of the available medical imaging techniques and to reduce invasiveness.

Currently, the most sophisticated techniques can image a cluster of tumor cells with a characteristic size of 5 to 10 mm—that is, more than 1 million cells. The ambitious goal is to produce devices that can identify single tumor cells, so even the last abnormal cell could be detected by in-vivo techniques. Here, the attention of our group is focused on the employment of targeted nanoparticles and the development of multiscale material models to differentiate types of tissues by measuring their biological and physical properties.

Different from materials that engineers are used to dealing with, biological tissues involve physical domains too irregular to be addressed by continuum theories, and too varied for a pure discrete approach at the molecular or atomic scale. Current mechanical theories cannot accurately address the modeling of materials as heterogeneous as biological matter.

Following original work by Vladimir Granik at the University of California, Berkeley, in the 1990s, our group is developing mathematical models based on the linear theory of doublet mechanics to accurately analyze the response of biological tissues to ultrasound-induced mechanical stresses. The theory of doublet mechanics, or DM, is a multiscale theory fully compatible with the continuum mechanics framework at the macroscale and with lattice dynamics at the atomic scale.

Different from the continuum theories, such as the theory of elasticity, doublet mechanics introduces scale factors in the formulation to capture the discrete and heterogeneous nature of biological materials. The DM theory has been combined with ultrasound-based non-destructive evaluation techniques leading to a new, non-imaging, tissue screening modality known as characterization-mode ultrasound (CMUS). The DM analysis of the spectral responses of normal and malignant tissues bombarded with ultrasound has shown significant and consistent differences, much larger than those measurable with a classical continuum analysis of spectral responses.

Researchers are working to develop nanoparticles smaller than blood cells to identify biomarkers and diagnose disease earlier than is possible today.

Results suggest the practicability and reliability of the method for biological tissue testing. We are currently developing such a system to be used as a pathologic evaluation tool providing a rapid, quantitative screening of biopsy specimens. More sophisticated models accounting for the multilayered structure of tissues such as the skin are needed to extend even further the field of application of CMUS.

The design and development of devices at the micro- and nanoscale for the early detection, therapy, and imaging of diseases require the convergence of knowledge and expertise pertaining to different fields, such as engineering, biology, and medicine. Functional design, structural analysis and optimization, fluid dynamics, and material selection are as important as physiology and pathology, biology and biochemistry, oncology, cardiology, and medical imaging. A transdisciplinary strategy should be initiated to support and catalyze synergistic research and education in this emerging area.

The trend toward transdisciplinarity is being reflected worldwide by science policy decisions, at least over the last two years. In 2004, the U.S. National Cancer Institute launched a $144.3 million five-year initiative, the Cancer Nanotechnology Plan, with the ambitious goal of eliminating death and suffering from cancer by 2015. Similar ideas were expessed by the European Union in 2005 with "Vision Paper and Basis for a Strategic Research Agenda for Nanomedicine." The rapid and robust spreading of new knowledge is fostered by the foundations of several new international journals, conferences, and associations. Even universities are creating new undergraduate and graduate programs in transdisciplinary studies bridging engineering and the biomedical and life sciences.

After decades of hyperspecialization, key advances with possible huge impacts on society can be envisioned only through cooperative research and education, which has to dissolve the old traditional boundaries among disciplines, institutions, and people. The integration of the fundamental concepts of continuum mechanics with the understanding of the small-scale weak interactions—such as van der Waals, electrostatic, and intermolecular forces—and with basic biomedical principles will open the doors to a new transdisciplinary field of research and education: bionanomechanics.

Cantilevers, Wires and Particles

Very small-scale diagnostic devices have been designed depending on the transduction mechanism—mechanical or electrical—of the biomedical stimulus. Devices that use nanocantilevers, nanowires, and nanoparticles have been built and tested, and are currently under development.

Nanowire-based devices: The sensors are nanometer-wide semiconductor wires coated with molecules (antibodies or fragments of antibodies) arranged in parallel on the bottom of a microfluidic chamber where a blood sample is introduced for analysis. The binding event among the molecules on the wire and the biomarkers in the sample produces a change in the electrical conductivity of the wire that can be measured in real time and related to the amount of biomarkers in solution.

Nanocantilever-based devices: The sensors are silicon cantilever beams a few hundred micrometers long, a few tens of micrometers wide, and about 1 micrometer thick. The beams are coated with molecules (antibodies or fragments of antibodies) and assembled in arrays of 10 or more in a fluidic chamber where the blood sample to be analyzed is introduced. The binding event between the molecules on the beam surface and the biomarkers in the blood sample induce a static deflection of the cantilever and a shift in resonance frequency that can be related to the amount of biomarkers in solution.

Nanoparticle-based systems: It is expected that porous and non-porous particles with a molecule-coated surface can be used to collect biomarkers. The nanoparticles have a characteristic submicrometer size and consequently can be introduced into the bloodstream through an intravenous injection. This system has the advantage of an in-vivo collection of the biological information avoiding any blood sampling and treatment.

Local Delivery

Small-scale medical devices of the future, because of their size, can be implanted in the body or circulate through the bloodstream to target cells and deliver medication directly to tumors.

Implantable devices are conceived as fixed systems comprising a chamber containing drug molecules in the milliliter range, and a filter with a characteristic size of a nanometer to control the release rate of the molecules.

These devices can be implanted subcutaneously, subbucally, near the primary tumor mass, or where the drugs should be released, and allow for the controlled release of drugs over long periods, perhaps months, avoiding periodic infusions.

Circulating devices would be nanoparticles used as smart drug vehicles comprising an internal reservoir carrying the therapeutic agents and an external coating with targeted chemico-physical properties. Different from the implanted devices, nanoparticles for drug delivery are sufficiently small to be injected through the veins and navigate within the blood vessels. As artificial biomimetic leukocytes, the cells of the immune system that defend the body against infectious disease and foreign materials, the nanoparticles circulate dragged by the blood flow searching for their biological target, the abnormal cell.

Paolo Decuzzi is associate professor of mechanical and biomedical engineering at the Center of Bio-Nanotechnology and Bio-Engineering for Medicine in the School of Medicine of the University of Magna Graecia in Catanzaro, Italy, and at the Center of Excellence in Computational Mechanics in the School of Engineering of the Politecnico di Bari. Mauro Ferrari is professor in the Brown Institute of Molecular Medicine at the University of Texas Health Science Center in Houston, professor of experimental therapeutics at M.D. Anderson Cancer Center, and professor of biomedical engineering at Rice University.