How to make bone…

I couple of weeks ago, my first proper-real-scientist article was accepted for publication in Acta Biomaterialia. This represents almost 2 years of work, so I was pretty pleased! Here, I’m going to try and explain exactly what I did in plain English, its not a very complex paper so this shouldn’t be too hard to follow!

You can read the full article here if you’re really that keen…

Some Background

The basic premiss for this paper is that current orthopaedic bone implants aren’t all that great. The first problem is something called a ‘modulus miss-match’. Bone itself is relatively flexible, but most implants are made out of titanium, which is pretty stiff. When we walk, our femurs compress under the strain, and this allows bone remodelling and strengthening. If we have a metal rod stuck into our femurs, however, the titanium will take all the compression and bone wastage will be the result. As a consequence of this problem, lots of current research is going into next generation bone implants with a modulus closer to that of bone.

The second problem with current generation implants is that cells just don’t like titanium surfaces that much. Stem cells exist in niches in the body. These niches allow stem cells the right environment to maintain their pluripotency and help them avoid genotoxic stress from their environment (protect them from cancer causing mutagens etc).

When a titanium rod is rammed into someones femur, these delicate niches are disrupted and the stem cells are forced out into the big bad world. These stem cells then take cues on how to differentiate from their surroundings. When, for example, bone stem cell’s are cultured in a large star shaped wells, they form massive contacts, spread out and put the cytoskeleton under immense strain. This causes the formation of osteoblasts (bone cells). However, when these pre-bone type cells are cultured in small rounded wells, they don’t spread out, form small transient surface adhesions and differentiate into adipose (fat) cells. This experiment was demonstrated rather eloquently by Kilian et al. in 2010.

The problem with titanium or ceramic hip implants is that they do nothing to direct cell differentiation. As these implants are not bioactive, they are treated by the body as foreign and encapsulated in a layer of adipose cells and general ‘mush’! This leads to a non-secure fit between the implant and bone tissue, and micro-motion of the implant itself. Basically the implant will shake when you walk and lead to bone erosion around the bone-implant barrier. This is why secondary hip replacement surgery is so much more complicated and why surgeons are reluctant to give primary replacements to young people – they know that the after 15 years the joint will be completely destroyed.

So in addition to designing next generation implants with less modulus miss-matching, we must design implants with surfaces capable of directing stem cell differentiation toward osteoblast lineage. This paper looked at how different nano-scale groove dimensions would direct ‘pre-osteoblast’ cell differentiation.

Osteoblast cell stained for vinculin (major focal adhesion protein) in green, actin (cytoskeleton) in red and nucleus in blue

Osteoblast cell stained for vinculin (major focal adhesion protein) in green, actin (cytoskeleton) in red and nucleus in blue

Results and Discussion

We used the common polymer, PCL, to culture our cells. This has been approved for use in vivo by the FDA and is biodegradable. The hope is that one day implants made from PCL may direct bone differentiation in stem cells then slowly degrade as the natural bone takes over. We embossed grooved patterns from glass slides (provided by Kelvin Nanotechnology) onto molten PCL and cultured primary human osteoprogenitors (in between stem cells and bone cells) to see what happened.

Our first problem was that the cells refused to grow on the surfaces. This is pretty common in science, what you think will take 2 weeks turns into a 5 month project! We guessed that this was probably due to PCL being pretty hydrophobic, so we plasma treated the surfaces and checked that the grooves were still in tact by atomic force microscopy. Plasma treatment basically bombards the surface with high energy plasma particles, this will disrupt the functional groups on the surface of the protein, creating positively and negatively charged groups (which are more hydrophilic). Atomic force microscopy drags an atomic-scale tip over a nano-scale surface and uses an algorithm to reconstruct the shape of the surface from vibrations in the tip caused by surface features. You can get some pretty cool single cell resolution pictures from this technique – so instead of showing you the (pretty boring) surface pictures in Figure 3, I’ll show some of these…

AFM tip tracing the outside of a cell cultured on a flat surface

AFM tip tracing the outside of a cell cultured on a flat surface


Schematic of image reconstruction after AFM

Schematic of image reconstruction after AFM



Once we had ascertained that a 30 second plasma treatment time could sufficiently increase the PCLs hydrophilicity to support the growth of cells, without compromising groove integrity, the experiment could go ahead.

We wanted to look at both the focal adhesions the cells made with the surface and the bone specific genes they expressed. Focal adhesions are formed where cells make contact with the surface they’re growing on, and like hands grasping for purchase, they allow both cell movement and cytoskeletal tension. Leading on from the work of Kilian (discussed above) and others, we hypothesised that as cells made larger and stronger focal contacts they would be more inclined to grow into osteoblasts. To characterise focal adhesions we used a antibody directed against the focal adhesion protein vinculin, coupled to a florescent probe (a technique known as immunocytochemistry).

Focal adhesions truncated after culture on grooved surfaces

Focal adhesions truncated after culture on grooved surfaces


Focal adhesion analysis; focal adhesions are traced in photoshop and an 'adhesion mask' is exported to calculate adhesion lengths

Focal adhesion analysis; focal adhesions are traced in photoshop and an ‘adhesion mask’ is exported to calculate adhesion lengths

We found that when cells were cultured on planar surfaces, they spread out more and formed longer focal adhesions. Cells cultured on grooved substrates formed fewer mature and super mature adhesions, and were observed to follow contact guidance along the edge of the ridge. Cells were thus less spread out and more polarised along the length of the nanometric grooves.

We also looked at expression of some bone specific proteins and transcription factors. As cells of the same lineage by-and-large look pretty similar, we needed something to tell us for sure whether the cells were becoming bone or fat. So we extracted cells RNA after 3 days cultured and looked at gene expression by both regular RT-PCR and oligonucleotide arrays. We found that, in-line with our hypothesis, when cells were cultured on planar surfaces they were more likely to differentiate down a osteogenic pathway.

In the paper, we hypothesised that as focal adhesions grow on planar surfaces, more FAC molecules are recruited to the adhesion. This causes an up regulation of an intermediate signalling molecule in the nucleus (ERK 1/2), and more transcription of the Runx2 master osteogenic transcription factor.

Cells on planar surfaces are able to spread out more freely, they form longer focal adhesions as they grow larger, which allows them to couple to larger actin stress fibres. As more tension is put on the focal adhesion, they elongate further to handle the tension and so the cycle is repeated, leading the cells to express bone specific transcription factors and ultimately differentiate down an osteogenic lineage pathway.

Although this paper just represents a small part of an ever enlarging field of research, it did help increase our knowledge of what a next generation implant will need. Other research has looked at changing the surface chemistry of implants, introducing random nanoscale pits and troughs and even using gold nano particles to help influence differentiation. Much further research is required in this field – but at least now we know that if you want bone, don’t use grooves!


Full article here…

JW Cassidy, JN Roberts, CA Smith, M Robertson, K White, ROC Oreffo, MJ Biggs and MJ Dalby. Osteogenic Lineage Restriction by Osteoprogenitors Cultured on Nanometric Grooved Surfaces – the Role of Focal Adhesion Maturation. Acta Biomaterialia Accepted Nov 2013

Written by John


I’m a recent Pharmacology Graduate from Glasgow, currently working toward a PhD in Cancer Research from the University of Cambridge. My main research aims are to understand the clonal dynamics in breast cancer, and how they are altered by therapy.

One thought on “How to make bone…

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>