Journal Club Post: 14

Glia U87 cells protect neuronal SH-SY5Y cells from indirect effect of radiation by reducing oxidative stress and apoptosis


Saeed Y1, Xie B1, Xu J1, Rehman A1, Hong M1, Hong Q1, Deng Y2.

Acta Biochim Biophys Sin (Shanghai). 2015 Apr;47(4):250-7. doi: 10.1093/abbs/gmv004. Epub 2015 Feb 27.

Abstract: Indirect effect of radiation increase oxidative stress enhances expression of apoptotic proteins leading to neuronal cell death. Irradiated cell-conditioned medium induced reactive oxygens species levels to be reduced in co-cultures of SH-SY5Y and U87MG cells. In co-cultures superoxide dismutase and glutathione antioxidant levels were increased. Co-cultures also inhibited expression of apoptotic proteins in irradiated cell conditioned media. This suggests glia may protect neuron cells by reducing oxidative stress and apoptotic death.

Introduction: Reactive oxygen species are major culprits prompting indirect effects of radiation resulting in neurodegerneration. Neurons are susceptible to stress but glia provide functional support to cope with hostile conditions including oxidative stress in vivo. Astrocytes provide trophic support and modulate synaptic activity. While glia have been previously reported as more resistant to oxidative stress than neurons. Therefore are neuron-astrocyte interactions enough to cope with oxidative stress and apoptotic damage induced by the indirect effect of radiation. Irradiated cell culture medium-induced reactive oxygen species were reduced in neurons when co-cultured with glia protecting mitochondrial integrity and inhibiting apoptotic proteins that cause neurodegeneration.

Methods: To study the protective effect of glia on irradiated neurons co-cultures grown in trans-well system with 0.4um pores. They measured cell viability by MTT method, which is a colorimetric assay based on mitochondrial ability to reduce MTT to a soluble formazan product with a purple colour with an absorbance at 490nm. In addition, they measured mitochondrial membrane potential using cell permeable mito-specific fluorescent probe JC-1. While they measured reactive oxygen species, nitric oxide, superoxide dismutase and glutathione using ELISA kits which read at 450 nm.

Results: Irradiated cell culture media (ICCM) reduced neuronal viability. However, glia were more stable at higher doses of ICCM. This suggests neurons are more sensitive. There was a decreased number of TUNEL-positive neurones when neuron-glia co-cultures were exposed to ICCM compared to ICCM exposed neurons only. This indicates glia may relieve neuronal stress. Also found in co-cultures neuronal cells were relieved from ICCM-induced mitochondrial depolarization relative to just ICCM-treated neurons. Neurons-glia co-cultures when exposed to ICCM  exhibited a reduction in nitric oxide concentration and increased superoxide dismutase and glutathione concentration relative to neuronal cells only which may be due to glia producing antioxidants. Pro-apoptotic proteins increased in ICCM exposed neurons while these had reduced expression in co-cultures with U87MG cells. While anti-apoptotic proteins increased in co-cultures while expression decreased in ICCM exposed neuronal cells only. So high anti and low pro-apoptotic proteins in co-culture infers glia help relieve oxidative stress in neurons.

Discussion: Neurons supported with metabolic and trophic support from glia that may protect neurons from oxidative stress. U87 shown to protect SK-N-SH cells against neurodegerative assault. OS may elevate apoptotic protein expression similar to Bak and Noxa in SK-N-SH neurons exhibiting similar pathological features to Parkinson’s disease. Therefore oxidative stress may be critical for cytotoxicity in neurons and neurodegeneration. Furthermore, glia appear to be more resistant to oxidative stress and apoptotic damage compared to neurons. Glia more resistant to oxidative stress than neurons. However the exact role of glia in neuroprotection unclear. This study demonstrated ICCM-induced oxidative stress was reduced in neurons in co-cultures with glia suggesting glia help reduce oxidative stress.  Antioxidant enzymes produced by glia to protect neurons from oxidative stress. This study suggests glia may help neurons cope with cytotoxic effect of ICCM by reducing oxidative stress protecting neurons from apoptosis. Further research is required to understand the underlying mechanism of how glia protect neurons.


Journal Club Post: 13


Osteoporotic hip fractures consist of cracks or breaks in the upper part of the femur bone due to a force that would not cause a fracture in a healthy bone. It is the most common type of osteoporotic fractures, resulting in considerable patient morbidity and mortality. Current treatments to prevent these fractures comprise the use of wearable protective devices or pharmacological treatments. However, they are limited by the lack of patients’ compliance, side effects and long delays in restoring bone properties. Internal augmentation of the proximal femur injecting polymethylmethacrylate (PMMA) bone cement may strengthen the bone overcoming the aforementioned limitations. Studies in the past have investigated femoral augmentation using large amounts of bone cement (40-50ml) and obtaining significant improvements in the femur yield and fracture load. However, injecting large volumes of PMMA into the bone may involve important risks such as thermal necrosis, embolism due to bone cement leakage or adjacent bone fracture. For this reason, some recent studies have decreased the amount of injected cement to 8-15ml. Outcomes show a high variability of the resulting femur strength depending on the location of the bone cement.


Reason for choice

The approach presented in this paper to design a generalized augmentation strategy for femoroplasty is innovative and relevant to my research area.



  1. Determine the ideal volume and distribution of bone cement to inject in a femoroplasty using numerical bone remodelling simulations.
  2. Develop new augmentation strategies and apply them to fifteen different FE models of femora.
  3. Compare the efficacy of the new augmentation techniques with a classical approach.
  4. Evaluate the effect of bone cement volume and placement.



  • Four new augmentation strategies were proposed, in which the cement distribution was idealized with cylindrical shapes. The main difference between the four versions was the alignment of the cylinder with respect to the medial and lateral ends of the proximal femur.
  • Stiffness, maximum load, yield load and yield energy were the studied biomechanical properties. Their increase was not significantly different amongst the new augmentation strategies, but there were relevant differences between the proposed strategies and the classical “single central” approach.
  • Augmentation was performed with approximately 12ml of PMMA in both the classical configuration and in one of the new proposed strategies. This resulted in a 2% and 11% increase in stiffness, 22% and 64% increase in yield load, 54% and 156% increase in yield energy, and 44% and 59% increase in maximum load, on average, respectively and compared to the non-augmented bone.
  • The relative increase of the biomechanical properties depended on the non-augmented properties of each femur. Hence, weaker bones experienced larger increases after augmentation than strong bones.



  • The main limitation of this research is the lack of experimental validation, although the results obtained with the “single central” approach were compared and found to be in agreement with previous experimental studies
  • Only the typical sideways fall load case has been studied using FEA. Examination of different load cases could provide more complete outcomes.
  • The idealization of the cement distribution with a cylindrical shape is an assumption that might not be possible to reproduce in clinical practice.
  • The bone cement polymerization process has not been addressed in this study. Therefore, the peak temperature reached during the femoroplasty remains unknown, as well as the potential debonding between the bone and the PMMA.



María Santana, Kingston University London



Varga, P., Inzana, J.A., Schwiedrzik, J., Zysset, P.K., Gueorguiev, B., Blauth, M. and Windolf, M., 2017. New approaches for cement-based prophylactic augmentation of the osteoporotic proximal femur provide enhanced reinforcement as predicted by non-linear finite element simulations. Clinical Biomechanics, 44, pp. 7-13.

Journal Club Post: 12


As the liver is the primary site of metabolism and a key target of drug-induced injury, a substantial amount of computer-based models which simulate pharmacokinetic and pharmacodynamic processes in the liver have been developed over the last decades. Models using in vitro data are increasingly being applied by the pharmaceutical industry to estimate in vivo pharmacokinetic parameters. Human or rodent hepatocytes and rodent liver microsomes widely serve as source to generate kinetic parameters related to metabolite formation (Vmax and Km) or drug substrate depletion (clearance or half-life). To estimate hepatic clearance (CLh), mathematical models may be used, such as

  1. the “well-stirred” model


where Qh = hepatic blood flow, fu = free fraction of drug in blood (or plasma), and CLint = intrinsic clearance

2. the parallel tube model

2    3. the dispersion model


The “well-stirred” model assumes that the liver is a single compartment in which a compound is instantaneously and homogenously distributed. Also, the free concentration of the compound in the liver is in equilibrium with and identical to the free concentration in the blood or plasma. This model is most often used due to it mathematical simplicity rather than any superiority over the other two. The parallel tube model assumes that the liver is organized as a succession of parallel tubes around which metabolic enzymes are evenly distributed with declining concentration along the length of each tube. The dispersion model is based on differential equations established on the assumption that the liver is cylinder-shaped and stems from the nonideal flow longitudinal dispersion model.

In order to apply in vitro data which are reported in μl/min per 106 cells or μl/min per mg microsomal protein to these models, in vitro data are scaled to up to what is estimated to be achieved in the in vivo assay by conversion to ml/min per standard rat/mouse weight.


Reason for Choice

This comparison between three established models to predict in vivo hepatic clearance (CLh), intrinsic clearance (CLint) and hepatic availability (Fh) provides me with a good foundation to build a model that estimates renal clearance. In contrast to other papers, this investigation uses a higher number of substances to establish a more robust comparison.



  1. To calculate in vivo CLint values from rat in vitro CLint data obtained from studies on isolated rat hepatocytes (35 drugs) and rat liver microsomes (52 drugs) using scaling factors.
  2. To afford a comprehensive comparison of the aforementioned liver models (“well-stirred”, parallel tube, and dispersion) to
    1. predict in vivo CLh and Fh from the in vivo CLint values calculated, and
    2. predict in vivo CLint values from experimentally generated in vivo CLh values,

based on the statistical parameters average fold-error (afe), root mean squared prediction error (rmse), and the correlation coefficient squared (r2).



In general, predictions from hepatocyte data are more precise (lower rmse) and less biased (lower afe) with higher r2 values than from the microsomal data.

For the microsomal prediction, the parallel tube and dispersion models show less bias than the “well-stirred” model but the precision was similar for all three models.

For the hepatocyte data, there was little difference in the bias between the models but more precision in the parallel tube and dispersion models relative to the “well-stirred” model.

The CLint predictions based on the parallel tube and dispersion models are similar to each other and more precise and less biased than the results obtained using the “well-stirred” model. Considering the mathematical complexity of the dispersion model, the parallel tube model is recommended to be used for the prediction of in vivo CLint.

The difference in the statistical estimates of bias and precision for the CLh prediction from all three models were only marginal. Therefore, the common practice of using the “well-stirred” model appears to be satisfactory.

Fh was overpredicted to a greater extent by the “well-stirred” model in comparison to the other two models.



What readers who are not familiar with such models might miss is the conceptual background of each of the models used.

The statistical analysis and graphical representations of prediction results are extensive and clearly outlined.

The brief discussion of potential reasons for why models based on microsomal data predict more poorly than models based on hepatocyte data provides a valuable insight for modellers to understand potential technical issues related to in vitro testing.



Julia Pletz, Liverpool John Moores University



Ito, K., and Houston, J.B. (2004) Comparison of the Use of Liver Models for Predicting Drug Clearance Using In Vitro Kinetic Data from Hepatic Microsomes and Isolated Hepatocytes. Pharmaceutical Research, Vol. 21, No. 5, pp. 785-792.

Also reviewed for additional information on the three clearance models: Boroujerdi, M. (2015) Pharmacokinetics and Toxicokinetics. CRC Press, Taylor & Francis Group, Boca Raton, FL, USA.

Henry Solomon (1853-1936)

Most recently as part of the DTA scheme we had a tour of the Wellcome Trust collection centre in Euston, London. Following on from this we thought it might be appropriate to write a short insight into the founder of the Wellcome trust, Henry Wellcome (1853-1936) and his contributions to science.


Henry was born in Wisconsin in 1853 and grew up as part of a farming family. Unfortunately, his father went bankrupt in 1861 after his crops had failed. Following this, the family headed to the Garden City. One year after the family moved to the Garden City, the land was invaded by a form of tribe called Sioux. Growing up Henry began to lead a group of children and they started to melt lead and cast rifle bullets for the defending settlers. He also helped his uncle in caring for the wounded.


Overtime, Wellcome’s father turned to god and became a second Adventist preacher and Henry himself worked for his uncle who ran a surgery and drug store.


In 1871, Wellcome went to study at the Philadelphia College of Pharmacy. This was an era in which qualified doctors had first started to use drugs to administer to patients. Here, Wellcome got an insight into the potential profitable  business that could be made out of pharmaceuticals. During his time at college he befriended, Silas Burroughs. They both moved to the UK in 1880 where they founded a pharmaceutical company which soon became a multinational enterprise. Henry became a leading figure in the British pharmaceutical industry, he established two laboratories one physiological and one chemical. He used them to prove that pharmaceutical research could help lead to better medicines. Some of the leading findings that Wellcome and his scientist found together were antitoxins for tetanus, diphtheria and gas gangrene. They also isolated histamine which lead to antihistamine production, standardizes insulin medicine.


Burroughs died in 1895 and Wellcome took over the company by himself. He began marketing pharmaceuticals in a very imaginative way that had not been done before at that time. During his life time Wellcome became fascinated with art and used to hold a series of lavish parties inviting ‘celebrities’ of that time.


In 1936 Henry died but left behind some what of a legacy. In his will he left a library, a trust, a collection and numerous other organistations. The Wellcome trust was the only shareholder of the Wellcome foundation. Over the following years shares were sold and eventually in 2000 Glaxo Wellcome plc merged with SmithKlineBeecham to create GalxoSmithKline. In 2004, the trust moved to its Euston headquarters where it acts as a free visitor centre, a place to learn, work and browse the collections that Henry developed during his lifetime.


Today the welcome trust provides large amounts of funding to various science projects around the globe. The welcome trusts’ policy is that it knows “good health makes life better” therefore they fell that funding they provide can help to improve health and help the best ideas to thrive. Some specialist areas the trust supports are; Diversity and inclusion, drug resistant infections, the planet and human health, research ecosysytems in Africa and Asia, science education and vaccines. The website is widely extensive and provides numerous schemes where all opportunities of research can be found. Wellcome funding supports over 14,00 people in more than 70 countries. Over the next five years the trust aims to spend up to £5bn helping thousands of researchers around the world to explore their scientific ideas, innovations, humanities and engagement with the public.


Any funding received from this prestigious trust is an honor and a highly recognised achievement.



Journal Club Post:11

The Impact of Phonological Versus Semantic Repetition Training on Generalisation in Chronic Stroke Aphasia Reflects Differences in Dorsal Pathway Connectivity



Language processing is supported by two independent, yet interactive, cortico-cortical pathways. The ventral pathway projects bilaterally to the ventrolateral prefrontal cortex, allowing interhemispheric connectivity of parallel-distributed processes that support auditory-articulation mapping via semantics. The dorsal pathway on the other hand is strongly left-lateralised and projects to the premotor cortex, where auditory-articulation mapping occurs in the absence of top-down semantic input, supporting both linguistic and non-linguistic auditory-motor processes. Dorsal and ventral pathway damage, such as that incurred through stroke, can result in speech and language impairment (aphasia). Dorsal pathway lesions for example, may produce selective speech production deficits, while sparing language comprehension; the opposite dissociation can be seen from ventral stream damage in that comprehension is impaired, yet production remains intact. Two different types of speech and language therapy have been used to rehabilitate these different types of aphasia; phonological therapies focus on rehabilitating speech production using phonology (spoken words accompanied a video of the mouth speaking the same word), while semantic therapies focus on rebuilding speech through semantic association (spoken words accompanied by a picture denoting the meaning of the word). It is argued that phonological therapies offer more rapid, but, short-term improvements, while semantic therapies can provide more beneficial therapeutic outcomes in the long-term, by enabling greater generalisability of language relearning to untreated stimuli. In cases of localised damage, however, it is debated whether the therapeutic focus should remain on restoring function in the damaged pathway, or whether therapy should attempt to compensate damage by utilising functioning of the spared pathway. To date, there has been insufficient research attention dedicated to exploring the interaction between brain damage severity in the dorsal and ventral pathways, and different approaches to speech and language rehabilitation in terms of therapeutic outcomes.

Reason for Choice

The findings of this paper may have implications for my PhD research in terms of how neuropsychological assessment and postoperative rehabilitation training in neurosurgical patients can be ‘tailored’ according to the location and severity of their specific surgical lesions.


  1. To determine whether patients with dorsal pathway damage would exhibit greater therapeutic gains and generalisability of relearning from repetition training via: (1) a restitutive approach engaging the damaged dorsal pathway (phonological therapy), or (2) a compensatory approach engaging the intact ventral pathway (semantic therapy).


  1. To examine how the severity of the patients’ brain damage interacts with training type. Specifically, the research aims to compare the therapeutic outcomes of restitutive and compensatory therapy in two individuals with chronic stroke aphasia (DM and JS), one with moderate brain damage and one with severe brain damage along the dorsal stream.


  1. To undertake neuroimaging to establish the extent of individual patient lesions.


  • For patients DM and JS, both semantic and phonological training were comparably effective in that repetition accuracy increased immediately post-intervention in comparison to baseline performance; these gains were also sustained for both patients at one-week follow-up.
  • However, when it came to generalisability to untrained items individual differences were observable between patients for the different interventions. DM demonstrated significantly higher repetition accuracy for phonological training, immediately and at follow-up, compared to JS. In contrast, JS showed the opposite pattern in that the semantic intervention provided a (marginally) significant increase in repetition accuracy immediately post-training, although this effect fell well below statistical significance at one-week follow-up.
  • In addition, MRI confirmed damage to the dorsal stream in both patients, specifically to the arcuate fasciculus (a key language tract), although DM’S damage was mild, while JS’s damage was more severe, indicating differences in connectivity.
  • The results suggest a double dissociation in terms of generalisability for different training types in patients with different dorsal pathway lesion severity.
  • Restitutive training may be effective for patients with milder damage, while more severe brain injury may benefit from compensatory interventions.


  • The only criticisms of this research are those regarding reduced statistical power and the limited training and follow-up periods.
  • While this research has yielded insightful findings, it is nonetheless important to replicate this study within a larger sample of patients to see whether this dissociation remains at a statistically significant level.
  • Finally, the longitudinal effects of different speech and language interventions for these patient subgroups should be considered, and perhaps, whether extended periods of compensatory training could provide better outcomes for more severe patients.


Reviewer: Rhiannon MacKenzie-Phelan, Liverpool John Moores University


Holland, R., Johns, S. L., & Woollams, A. M. (2016). The impact of phonological versus semantic repetition training on generalisation in chronic stroke aphasia reflects differences in dorsal pathway connectivity. Neuropsychological Rehabilitation, 1-20. doi:10.1080/09602011.2016.1190384




Journal Club Post :10

Three-dimensional printing of complex biological structures by freeform reversible embedding of suspended hydrogels


Tissue engineering is growing evermore as a method for addressing the degradation/failure of tissues and organs. Hydrogels are excellent tools for tissue engineering because they contain a microstructure similar to that of extracellular matrix (ECM). They essentially provide a scaffold around which tissues can be engineered with an architecture that is reflective of the native tissue being targeted. Current approaches for the 3D printing of biological hydrogels have achieved important advances but are still in need of significant improvement. Biopolymer materials are not particularly conducive to 3D printing and suffer from poor resolution and hence difficulty in reproducing the desired printed shape. This is due to the low viscosity in the pre-gelled state causing them to flow and deform once deposited. A solution to this is to print ‘a gel within a gel’.

Reason for choice

The work in this paper holds many similarities to the work that I am carrying out in my PhD however our differing methods allow me to compare and contrast our end results and optimise the protocol.


  • Supporting soft structures as they are printed so that they do not collapse or deform
  • anisotropically depositing the material to match the microstructure of real tissue,
  • keeping cells alive during this whole process using aqueous environments that are pH-, ionic-, temperature-, and sterility-controlled within tight tolerances


untitledFig. 3 FRESH printing of biological structures based on 3D imaging data and functional analysis of the printed parts.(A) A model of a human femur from 3D CT imaging data is scaled down and processed into machine code for FRESH printing. (B) The femur is FRESH printed in alginate, and after removal from the support bath, it closely resembles the model and is easily handled. (D) A model of a section of a human right coronary arterial tree from 3D MRI is processed at full scale into machine code for FRESH printing. (E) An example of the arterial tree printed in alginate (black) and embedded in the gelatin slurry support bath. Scale bars, 4 mm (B) and 10 mm (E).


  • Construct sizes have to be scaled down to achieve desirable print resolution.
  • The temperature has to be raised to 37 °C after the construct has been printed at 22 °C in order to remove the supporting gelatin bath (I have found alternative methods that allow constructs to be printed and removed at any temperature (within reason)).
  • The resolution and morphology of a print depend on a number of machine settings and require optimization for each material used
  • Dimensional analysis comparing the 3D models to the 3D printed constructs demonstrated variability in length, width, and size of major internal structures (although somewhat low).
  • Higher resolution is possible using higher-precision printers however, commercially available and custom-built printers currently cost more than $100,000 and/or require specialized expertise to operate.


DOI: 10.1126/sciadv.1500758                                                                   Reviewed by: Jessica Senior, University of Huddersfield

Journal Club Post: 9

Diabetes and Dementia?!


Elderly people with type 2 Diabetes have double the risk of developing dementia compared to those without Diabetes The suggested cause of this is due to microvascular pathology in the brain. Abnormalities in the retinal microvasculature may therefore represent a proxy for cerebral microvascular disease. As the retinal vessels can be directly seen, the microvascular changes in the retina may offer a window into the microvasculature of the brain.


Although this isn’t related to my topic I thought it might be more readable for those not familiar with ocular conditions.


To determine whether severe diabetic retinal disease is associated with an increased incidence of dementia identified during 10 years of follow-up in a large well-characterised cohort of older patients with type 2 diabetes


  • Patients with diabetic retinal disease had a 42% increased risk of dementia compared to those without
  • They have therefore concluded that microvascular disease does contribute to the development of dementia when type 2 diabetes is present


A lot of the information collected about the patient’ lifestyle, diabetic history and height & weight was self-reported

Participants from lower socio-economic backgrounds were excluded so may be under reporting of diabetic retinopathy

American study therefore those who responded have medical insurance and can afford to have regular diabetic screening and monitoring done

Could it just be a coincidence? They had a sample size of approx.. 30,000 people – if 30,000 elderly people are monitored over 10 years for dementia, chances are some dementia signs will be found.

Claire Gorman

Plymouth University