Hereditary spastic paraplegia (HSP) is a group of neurodegenerative monogenic disorders characterized by dysfunction of the longer tracts of the spinal cord: the corticospinal tract and, to a lesser extent, the posterior columns[1 ],[2 ],[3 ],[4 ]. This culminates in the key clinical aspects of HSP: spasticity of the lower limbs, followed by a - usually less severe - weakness, urinary urgency and decreased vibration sense[1 ],[2 ],[3 ],[4 ]. The HSP may segregate as autosomal dominant (AD-HSP), autosomal recessive (AR-HSP), X-linked or mitochondrial traits[1 ],[2 ],[3 ],[4 ]. On clinical grounds, one may present with a pure form of the disease, which is more frequent in families with an autosomal dominant inheritance pattern. On the other hand, a complex or complicated clinical picture, characterized by the presence of other neurological or non-neurological dysfunctions, is more frequently found in AR-HSP or X-linked HSP[1 ],[2 ],[3 ],[4 ]. These features encompass, for example, ataxia, parkinsonism, peripheral neuropathy, cognitive dysfunction, cataracts, and icthyosis[1 ],[2 ],[3 ],[4 ],[5 ],[6 ],[7 ].
The diagnosis of an inherited spastic paraplegia is based essentially on the clinical picture, together with the investigation of other, potentially affected, members of the kindred. Exclusion of acquired myelopathies is fundamental in the absence of family history. Some conditions that can mimic HSP are: human T-lymphotropic virus-related myelopathy, primary progressive multiple sclerosis, vitamin B12 deficiency, copper deficiency, spinal cord tumors or malformations and degenerative disorders of the spine[4 ],[5 ],[6 ],[7 ]. The presence of a family history or consanguinity strengthens the possibility of a hereditary condition; its absence, on the contrary, should not dissuade the clinician from the diagnosis[1 ],[2 ],[3 ],[4 ],[5 ],[6 ],[7 ]. An apparently isolated case can be related to the occurrence of mildly affected family members, incomplete penetrance, false paternity or de novo mutations, for example.
In the last three decades, advances in molecular genetics have revealed that HSPs constitute a large and heterogeneous group of diseases[4 ],[5 ]. The knowledge of the key historical marks of HSPs may assist neurologists in understanding why a common ground of spasticity and weakness of the lower limbs has evolved into more than 80 different genetic subtypes. These frequently overlap clinically and/or genetically with other degenerative disorders such as ataxias, leukodystrophies, spinal muscular atrophy, amyotrophic lateral sclerosis (ALS) and peripheral neuropathies[7 ]. Historically, the differentiation between HSPs and inherited ataxias has been the most challenging. The most common ataxia worldwide, SCA3, as well as other autosomal dominant ataxias (examples: SCA1, SCA2, SCA6, SCA7), are frequently accompanied by pyramidal signs, and spastic paraparesis can be the inaugural symptom[8 ],[9 ],[10 ]. In regard to the AR group, absent reflexes are a hallmark of its most frequent subtype, Friedreich's ataxia. The late form of Friedreich's ataxia, beginning at age 25 or older, is alternatively associated with retained reflexes, and may present initially as spastic paraparesis[11 ],[12 ]. The key to differentiating complicated HSP from ataxias with associated corticospinal tract dysfunction is to focus on the patient's main disabilities. As a rule, individuals with HSP are overtly spastic and have mild or no complaints relating to the upper limbs. Whereas, in the ataxia group, there is no pronounced dissociation between upper and lower limbs. Although useful on clinical grounds, more recently-discovered genes are increasingly challenging the neurologist's abilities to differentiate these conditions. An ever-growing number of genetic subtypes can present either with pure ataxia or pure spastic paraparesis, and a discordant phenotype can occur within the same family. Some examples of this overlap are: HSP-SPG7, autosomal recessive spastic ataxia of Charlevoix-Saguenay, HSP-SPG35, HSP-SPG46, cerebrotendinous xanthomatosis, Alexander's disease and spastic ataxias (SPAX 1 to 5)[13 ],[14 ],[15 ],[16 ].
Considering the complexity associated with the study of HSP, the present article is devoted to taking the reader through the most import discoveries in the field, from its original description in 1880, through to the challenges faced today. We have focused on the outstanding works regarding clinical, neuropathology, neuroimaging, neurophysiology, and genetics of these rare motor neuron disorders.
Hereditary spastic paraplegia - first descriptions
In 1880, the Baltic-German (region known today as Estonia and Latvia) neurologist Ernst Adolf Gustav Gottfried von Strümpell ([Figure 1 ]) published the first case series of patients with HSP. Strümpell reported on two siblings with a probable AD-HSP whose symptoms manifested at 37 and 56 years of age. Clinically, these patients showed a pure form of spastic paraplegia. After the death of one of the siblings, neuropathology showed degeneration of the lateral corticospinal tract, fasciculus gracilis and spinocerebellar tract[17 ],[18 ]. Wilhelm Erb was probably the first to describe a condition dominated by limb spasticity. Although controversy persists, most authors today consider Erb's works to be the first description of primary lateral sclerosis. A case description by Otto Adolph Seeligmuller, in 1876, has long been disputed as the first description of HSP. Nevertheless, a critical analysis of Seeligmuller's publication highlighted that the patient studied had a clinical picture dominated by muscular atrophy and bulbar paralysis, features not suggestive of HSP itself. In 1888, Maurice Lorrain, a French neurologist, published a more detailed contribution to the anatomical and clinical study of HSPs; hence HSP also being known as Strümpell-Lorrain disease[19 ],[20 ].
Figure 1 Ernst Adolf Gustav Gottfried von Strümpell (1853-1925).(Extracted from http://www.200.uk-erlangen.de/de/geschichte/20-koepfeder-erlangeruniversitaetsmedizin/struempell/, April 08th, 2017).
Clinical phenotype
Rhein[21 ] was the first, in 1916, to drive attention to the exquisite clinical heterogeneity some families with HSP displayed. His observations were accompanied by many case series describing pleomorphic clinical pictures encompassing: retinal degeneration[22 ],[23 ], dementia[24 ], extrapyramidal symptoms[25 ], mental handicap[26 ], hand amyotrophy[27 ] and other features. The term Strumpell's familial spastic paraplegia was reserved for families that displayed only spastic paraparesis, a condition that was thought to be much rarer[28 ]. Currently, it is recognized that pure forms are more prevalent than complicated ones, their relatively benign clinical course, not affecting life expectancy, greatly contributed to their underdiagnosis at that time. Most epidemiological studies were based on postmortem or hospitalization records[29 ].
Anita Harding, a professor of neurology at the University of London, was an important pioneer in the field of molecular neurogenetics. In the 1980s, she published a series of groundbreaking works addressing HSP. In 1981, she presented what was, at the time, the largest investigation into pure HSP, with 22 families studied[30 ]. Eighteen presented with AD-HSP, three were AR-HSP and, in the remaining kindred, the inheritance was uncertain due to the paucity of individuals affected. This work has consolidated the knowledge that pure subtypes of the disease are almost always dominantly inherited[30 ]. Harding also shed light on the importance of examining first degree relatives who considered themselves asymptomatic, as that practice unraveled AD-HSP in five of the 22 kindreds in this work[30 ]. She also reinforced the knowledge that spasticity, and not weakness, is the main source of disability in this scenario, an aspect that remains useful in differentiating HSP from other myelopathies[5 ],[31 ]. Harding's major contribution to the field of HSP was published in The Lancet journal in 1983[32 ]. Entitled “Classification of the hereditary ataxias and paraplegias”, this paper established an accurate differentiation between ataxias and HSP, providing a workable base for the etiological investigations that would come[32 ]. The now current clinical classification of HSP into pure and complicated forms was proposed. Harding defined that, besides the spastic paraparesis, additional mild signs such as vibration and segmental position sense deficits, slight distal amyotrophy and sphincter dysfunction were also conceivably present in the pure phenotype. Before this work, the literature had been confusing with regard to this topic. Pure HSP was subdivided into two groups according to age of onset, before or after 35 years of age. The complicated forms were recognized to be rarer and genetically heterogeneous. This work also provided refinement of the HSP phenotype, emphasizing the variability of disease progression, even within the same family. Sadly, Harding's outstanding career ended prematurely at the age 42 when she succumbed to a cancer[33 ]. Professor Victor Dubowitz wrote, on the occasion of her obituary: “the neuromuscular world and, indeed, world neurology has lost one of its most colorful, most productive, and best-loved characters”[33 ].
In 2006, the German Network for Hereditary Spastic Paraplegia presented the spastic paraplegia rating scale, developed to quantify the disease progression clinically. Validated measures of disease severity are essential to understanding the disease's natural history. Additionally, by their ability to measure treatment impact, they are decisive in the development of future clinical trials[34 ].
Structural characterization: neuropathology and neuroimaging
While the clinical descriptions of HSP flourished during the 1980s, the pathological comprehension of the disease saw flourishing moments 100 years before. We return to Adolf von Strümpell, the first to identify the disease clinically, who was also the first to describe its pathological hallmarks six years later, in 1886[35 ]. An unrelated case was also documented by him in 1904[36 ]. Both pathological specimens showed degeneration of the corticospinal tract in the spinal cord with a distal-proximal gradient. Since then, the disease has been understood as a distal axonopathy of the longest large myelinated fibers of the spinal tract. In 1952, Schwarz[37 ]gave an extensive review of the pathologic literature, highlighting the contributions of Newmark (1904, 1906, 1911)[38 ], Jakob[39 ] and Kahlstorf[40 ], that culminated with the differentiation of Strümpell's disease from other conditions such as cerebellar ataxias and motor neuron disorders, establishing that, at first, lesions are restricted to the spinal cord and affecting the corticospinal and posterior tracts[41 ]. In 2004, DeLuca and colleagues quantitatively examined the neuronal population of six HSP patients, the largest pathological report of the disease[42 ]. Marked symmetrical reduction in area and axonal density of the corticospinal tract in the lower regions of the spinal cord was found. On the other hand, sensory tracts demonstrated a significant reduction in area and axonal density only in the upper regions of the spinal cord. This supported the concept that the extent of axonal loss in HSP is tract-specific[42 ].
In the past decade, neuroimaging techniques have emerged as powerful tools to investigate structural abnormalities in a wide variety of neurological disorders in vivo. In this field, Agosta et al. [43 ], were able to investigate a larger cohort. They demonstrated that microstructural brain abnormalities are rather diffuse and not restricted to motor pathways, when one analyzes multiple subtypes of HSP. Lindig[44 ] and Rezende[45 ], both in 2015, made substantial contributions to the delineation of the extent of neurodegeneration specifically associated with HSP- SPG4. Both authors showed correlations between white matter tract disruption and disease severity, indicating that diffusion tensor imaging measures could be used as biomarkers. Rezende and colleagues also demonstrated the absence of cortical thinning as a hallmark of this pure subtype of HSP[44 ],[45 ]. Regarding HSP-SPG11, both França et al.[46 ] and Pan et al.[47 ] revealed widespread white matter microstructural disruption. Pan et al. also demonstrated a distal-proximal gradient, whereas França et al. identified major deep grey matter volumetric loss[46 ],[47 ].
Neurophysiology
At the same time that Anita Harding made her contributions to the HSP classification, Schady and colleagues[48 ] published their neurophysiological findings of central motorconduction studies by transcranial magnetic stimulation in HSP. The abnormalities found were worse in the lower than in the upper limbs, confirming neurological examination findings and strengthening the hypothesis made from observing pathological specimens in a length-dependent dysfunction[48 ].
Advances in genetics
After the accurate clinical classification developed by Harding[32 ], the research involving HSP evolved from anecdotal case descriptions to the systematic study of families alike. Numerous HSP genetic studies have been published since the 1980s. Several genetic subtypes of HSP have been described and numbered sequentially, based on the order of the gene discovery (SPG1, SPG2, SPG3, etc. j[1 ].
In 1994, Jouet et al.[49 ], reported that three disorders; X-linked spastic paraplegia, MASA syndrome (mental retardation, aphasia, shuffling gait and adducted thumbs) and X-linked hydrocephalus were, in fact, allelic conditions, all resulting from mutations in the gene for neural cell adhesion molecule (L1CAM or SPG1 )[49 ]. This was the first gene discovered that caused HSP; L1CAM codes for an axonal glycoprotein involved in neuronal migration and differentiation. Also in 1994, Hazan et al., identified the locus implicated in the majority of cases of dominantly inherited HSP. The SPAST or SPG4 gene was cloned in 1999 by the same group and responds to up to 60% of AD-HSP[50 ]. This gene encodes spastin, an AAA protein family member related to microtubule severing and endoplasmatic reticulum morphogenesis, key factors in the process of intra-axonal transport[50 ].
In 1998, Griffiths and colleagues were the firsts to create an animal model of HSP[51 ]. They developed a mouse knockdown of proteolipoprotein (causing HSP-SPG2), unraveling the proteins critical role in maintaining the integrity of myelinated axons. Affected mice developed a clear spastic gait, related to fiber swelling and impaired axonal transport[51 ]. This model opened up a new window of understanding of the mechanisms through which corticospinal axons degenerate in HSP. During the same year, the first gene for autosomal recessive HSP was identified[52 ], when the HSP-SPG7 gene was cloned. Paraplegin, encoded by the SPG7 gene, localizes to the mitochondria, and ragged red muscle fibers (a typical finding in mitochondrial disease), and was readily detected in this work. Key findings suggestive of this phenotype encompass ataxia, ptosis, progressive external ophthalmoplegia and optic atrophy. The latter can also occur in isolation as an autosomal dominant trait[52 ].
In 2007, SPG11 was identified as the gene most commonly related to spastic paraplegia with mental impairment and thin corpus callosum. Stevanin, led by Brice and other colleagues, identified loss of function mutations affecting this gene in 11 families[53 ]. Since then, HSP-SPG11 has been recognized as the most common AR-HSP. Surprisingly, SPG11 mutations can also primarily present with other phenotypes, such as parkinsonism or ALS[54 ].
The next-generation sequencing techniques, encompassing gene panels, whole exome sequencing and whole genome sequencing have initiated a new era of knowledge in the field. The possibility of sequencing dozens, or hundreds, of genes simultaneously, has enormously increased the number of genes known and has expanded both the phenotype and genotype of HSP[54 ],[55 ].
A brilliant example of the new possibilities opened up by this technology is the work done by Gaia Novarino et al., an Italian physician devoted to unraveling the molecular pathways of many neurological disorders[55 ]. Novarino's work, published in Science in 2014, presented the discovery, and functional validation of, as many as 18 new HSP-related genes. Her investigations have also contributed to linking HSP-related genes to other degenerative disorders, especially ALS, Alzheimer's and Parkinson's diseases. Today, more than 80 different loci related to hereditary spastic paraplegias are recognized, with more than 50 genes already identified[56 ].
Brazilian contributions
Brazilian researchers have made substantial contributions to the field of HSP. Clinically, Teive et al.[57 ] described the first cases of HSP with thin corpus callosum in Brazil in 2001, at a time when very few families worldwide had been identified, other than in Japan[57 ]. In 2007, a multinational group of researchers, including the Brazilian biologist Zatz, identified mutations causing HSP-SPG8[58 ]. In 2014, França et al.[59 ] addressed the SPG4 molecular epidemiology, confirming it as the most common HSP in Brazil[59 ]. In 2015, Melo et al.[60 ] described the over-expression of the KLC2 gene as the cause of spastic paraplegia with optic atrophy and neuropathy. This discovery was unique in the field, as it was the first HSP-related mutation to be documented in an untranslated region[60 ].
